working paper series - sec.gov | home · that eased restrictions under sor would result in firms...
TRANSCRIPT
Working Paper Series
Differences in the Information Environment Prior to SEOs under Relaxed Disclosure Regulation
Sarah B Clinton Joshua T White Tracie Woidtke
DERA Working Paper 2013-05
NOTE Staff working papers in the DERA Working Paper Series are preliminary materials circulated to stimulate discussion and critical comment References in publications to the DERA Working Paper Series (other than acknowledgement) should be cleared with the author(s) in light of the tentative character of these papers The Securities and Exchange Commission as a matter of policy disclaims responsibility for an y private publication or statement by any of its employees The views expressed herein are those of the author and do not necessarily reflect the views of the Commission or of the authorrsquos colleagues on the staff of the Commission
Differences in the information environment prior to seasoned equity offerings under
relaxed disclosure regulation
Sarah B Clinton
Department of Accounting and Information Management
College of Business Administration
University of Tennessee
620 Stokely Management Center Knoxville TN 37996
clintonutkedu
Joshua T White
US Securities and Exchange Commission Division of Economic and Risk Analysis
Office of Corporate Finance 100 F Street NE Washington DC 20549
whitejosecgov
Tracie Woidtke
Department of Finance
College of Business Administration
University of Tennessee
425 Stokely Management Center Knoxville TN 37996
twoidtkeutkedu
Telephone (865) 974-1718
Fax (865) 974-1716
April 28 2014
We thank an anonymous referee SP Kothari (editor) Anup Agrawal Cindy Alexander Judy
Beckman Bruce Behn James Chyz Shayne Clinton Donial Dastgir Raquel Fox Joan
Heminway Gerard Hoberg Rebekah McCarty Sarah McVay Alex Nekrasov Marlene Plumlee
Andy Puckett Anup Srivastava Amy Starr Karen Ubell Christian Vossler and seminar
participants at the 2010 American Accounting Association Northeast regional meeting 2011
American Accounting Association annual meeting 2012 Financial Accounting and Reporting
Section midyear meeting Southern Methodist University Texas AampM University University of
Alabama University of Arizona Clemson University Miami University University of
Delaware University of Georgia University of Mississippi University of NebraskandashLincoln
University of Utah US Securities and Exchange Commission (SEC) and the University of
Tennessee Corporate Governance Center for insightful comments and suggestions As a matter
of policy the Securities and Exchange Commission disclaims responsibility for any private
publication or statement by any of its employees Although this paper was reviewed by other
staff members within the Commission the views expressed herein are those of the authors and
do not necessarily reflect those of the Commission the Commissioners or other members of the
Staff
Corresponding author
Differences in the information environment prior to seasoned equity offerings under
relaxed disclosure regulation
ABSTRACT
The Securities Offering Reform (SOR) promulgated by the US Securities and Exchange
Commission (SEC) in 2005 represents a major change in the seasoned equity offering (SEO)
process SOR eases disclosure restrictions and reduces uncertainty regarding disclosures allowed
prior to an SEO The SEC argued that SOR would result in an improved information
environment and benefit capital formation efficiency In contrast critics claimed that SOR would
allow firms to hype their stock before an SEO to the detriment of the information environment It
is also possible that SOR is not effective in generating greater disclosure and affecting the
information environment during equity capital formation because of ongoing concerns such as
the litigious climate in the US This paper is the first to examine differences in disclosure and
the information environment at the time of seasoned equity capital formation under SOR
Overall the results suggest that relaxed disclosure restrictions under SOR are associated with
greater disclosure immediately preceding the SEO issue date and greater disclosure is related to
a richer information environment with capital formation benefits During the month before the
SEO issue date under SOR we find more frequent disclosure of management earnings forecasts
and Form 8-K filings and management earnings forecasts are more accurate In addition we find
disclosure within a week before the SEO issue date under SOR is associated with greater
absolute market-adjusted returns (ie information magnitude) and more positive stock returns
with no reversal afterward (ie capital formation benefits through informative disclosure)
Keywords Securities Offering Reform Securities regulation Seasoned equity offerings
Disclosure Information environment Shelf registration
JEL classification M41 G38 K22
Differences in the information environment prior to seasoned equity offerings under
relaxed disclosure regulation
1 Introduction
Itrsquos a basic precept of American securities law that shareholders should be given the information
they need to evaluate their companies
- Robert Jackson professor of law Columbia University (New York Times April 24 2013)
The information environment plays a critical role in capital market efficiency However
the regulatory framework and the litigious climate within the US often result in firms disclosing
only information that is required (Cadwalader Wickersham amp Taft LLP 2005 Rogers and Van
Buskirk 2009) Until recently disclosure prior to seasoned equity offerings (SEOs) had been
limited by the US Securities and Exchange Commission (SEC) even though investor demand
for information was high Section 5(c) of the Securities Act restricted firm disclosure before an
SEO to ban any attempts to ldquocondition the marketrdquo The SEC promulgated the Securities
Offering Reform (SOR) in 2005 to modernize the SEO process and eliminate restrictions that
were no longer viewed as necessary to protect investors yet prevented the dissemination of
accurate information1
This paper investigates whether SOR is associated with differences in
disclosure and a richer information environment during the equity capital formation period (ie
the period preceding an SEO issue when investors assess the SEO price and decide whether to
invest)
SOR is a unique shift in the regulatory environment for SEOs In addition to removing
restrictions on factual and forward-looking disclosures prior to SEOs SOR provides a safe
harbor for disclosures and allows a novel disclosure venue the free writing prospectus (FWP)
which is any written communication that offers to sell securities that are or will be subject to a
registration statement and can include factual and forward-looking information Critics argued
1 See the Securities Offering Reform Final Rule at httpwwwsecgovrulesfinal33-8591frpdf
1
that eased restrictions under SOR would result in firms providing opportunistic disclosure to
ldquohyperdquo expectations before an SEO (eg Morrissey 2007) However the SEC argued that the
primary benefit of SOR was to provide greater and timelier information flow when investors
decide whether to invest in an SEO (SOR Final Rule p 44792) Despite the increased flexibility
under SOR greater disclosure during equity capital formation may not occur for several reasons
Firms may be concerned that greater disclosure could reveal proprietary information (Ali et al
2014) or establish a disclosure precedent that is difficult to maintain or deviates from a long-
standing policy (Graham et al 2005) Firms may also continue the pre-SOR policy of limiting
disclosure during the capital formation period due to liability concerns such as shareholder
litigation or SEC sanctions for conditioning the market (Cadwalader Wickersham amp Taft LLP
2005) Thus it is not clear that SOR automatically leads to greater disclosure or an improved
information environment prior to SEOs
We analyze a sample of 360 SEOs issued by well-known seasoned issuers (WKSIs) a
new class of issuers established by SOR from 2002 to 2009 To provide the first evidence of
whether disclosure is greater and the information environment is richer under SOR during the
capital formation period we conduct tests using multiple forms of factual and forward-looking
disclosure Specifically we examine management earnings forecasts (MEFs) FWPs 8-K filings
and earnings announcements To capture the capital formation period we focus on the month
(ie 30 days) prior to the SEO issue date The role of disclosure before the issue date is
important as a result of changes in the equity offering process including a shift to shelf
registrations and an increase in the distance between the SEO filing and issue dates under SOR
(Table 1) In addition to our overall analysis we separately examine firms with a policy of
providing MEFs as well as other disclosures (forecasting sample) and firms with a policy of
2
providing only disclosures other than MEFs (non-forecasting sample) to investigate whether
firms behave differently under relaxed disclosure regulation based on prior disclosure practices
We conduct three primary tests First we test whether SOR is associated with more
factual and forward-looking disclosure Second we test whether MEF disclosure is indicative of
a richer information environment under SOR for the forecasting sample We examine MEFs to
see if they are more accurate (ie informative) potentially reducing information asymmetry and
litigation risk or overly optimistic (ie opportunistic) as predicted by critics of SOR Third we
analyze stock prices in two ways to further evaluate the information environment and investigate
capital formation benefits We examine absolute market-adjusted returns to measure the
magnitude of information and signed raw returns to investigate whether disclosure is
opportunistic or informative
Overall the results suggest that relaxed disclosure regulation under SOR is associated
with greater disclosure prior to SEOs which is related to a richer information environment with
capital formation benefits Contrary to criticisms of SOR we do not find evidence that these
disclosures are opportunistic in nature The use of FWPs in the full sample and a greater
proportion of annual MEFs in the forecasting sample under SOR result in a 51 greater
proportion of SEOs preceded by guidance disclosure during the capital formation period The
overall frequency of disclosure is 25 greater and the amount of information provided in the 8-
K immediately before the issue date more than doubles in size Furthermore the number of days
between the last disclosure before an SEO and the issue date is lower under SOR Also MEFs
immediately preceding the SEO issue date are 68 more accurate under SOR (ie have reduced
error and bias) and tend to include more downward than upward revisions In addition the
abnormal magnitude of information reflected in absolute market-adjusted returns is almost twice
3
as large after SOR and the difference is driven by SEOs with disclosure within one week before
the issue date Furthermore stock returns during the capital formation period are greater after
SOR for SEOs with disclosure within one week prior to the issue date with no reversal after the
issue date consistent with enhanced capital formation through informative not opportunistic
disclosure These results support the SECrsquos argument that SOR would result in an improved
information environment and benefit capital formation efficiency
Our paper makes several contributions to the literature First we contribute to the
literature on disclosure and the information environment Specifically our paper is the first to
our knowledge to examine different forms of disclosure including the FWP during the equity
capital formation period Second we contribute to the SEO literature by highlighting changes in
the SEO environment that have important implications for future research For example we
show that the capital formation period is different from the pre-SEO period defined in earlier
studies for the majority of SEOs during our more recent sample period Although earlier studies
frequently designate the filing date as the SEO announcement date (eg Asquith and Mullins
1986 Masulis and Korwar 1986) because few SEOs were shelf-registered studies covering
more recent periods and using the filing date as the SEO announcement date may misinterpret
insignificant returns as less negative reactions to SEOs Less information is conveyed at the
filing date as a result of SOR Finally we contribute to research on regulatory reform While
other studies on relaxed regulation tend to find negative outcomes our results indicate that SOR
increases information flow and smoothes capital formation2
Our paper advances the literature on
disclosure SEOs and regulatory reform and should be of interest to the SEC and firms that
access capital on the secondary market
2 Empirical evidence of negative outcomes is linked to relaxed regulation in disclosure (eg Fernandes et al 2010)
and financial markets (eg Beck et al 2010)
4
The remainder of the paper is organized as follows Section 2 provides a discussion of
SOR and presents our empirical questions Section 3 describes the data and Section 4 outlines
our research design Section 5 presents our results Section 6 concludes the paper
2 Background and empirical questions
21 Before the Securities Offering Reform of 2005
Before SOR the SEO process was commonly divided into three periods pre-filing
waiting and post-effective (see Fig 1 for shelf registration timeline) The pre-filing period began
when a firm contemplated an SEO and ended at the filing date when the registration statement
was filed with the SEC The waiting period began at the filing date and ended when the SEC
declared the registration statement effective During the post-effective period a shelf-registered
SEO firm usually filed a preliminary prospectus (pre-issue date) before the shares were sold
(issue date) The post-effective period was typically short to nonexistent for non-shelf (ie
traditional) registrations because the filing date which was also the pre-issue date and issue date
were very close (Bethel and Krigman 2008)
[Insert Fig 1 near here]
Prior to the reform certain communication was prohibited in the pre-filing and waiting
periods and was limited in the post-effective period The pre-filing and waiting periods were
covered by disclosure restrictions banning any attempt to condition the market Any additional
disclosures in connection with the SEO including forward-looking statements were expressly
forbidden with the exception of a prospectus During the post-effective period factual and
forward-looking disclosure was effectively limited through cumbersome prospectus amendment
requirements and liability concerns (Latham amp Watkins LLP 2005) For example forecasts had
5
to be filed as part of the prospectus Many issuers did not want to include a forecast in a
prospectus because it could be viewed as conditioning the market3
Any disclosure that could be
viewed as an offer (eg an MEF) was limited to statutory prospectus disclosure Furthermore
once the final prospectus was prepared issuers were responsible for ensuring that an investor had
received the final prospectus and the SEC reserved the right to review the final prospectus
which could potentially delay the SEO
22 After the Securities Offering Reform of 2005
SOR became effective December 1 2005 Although SOR affects all issuers it provides
the most significant changes for a newly established class of issuers WKSIs which have a
market capitalization of at least $700 million4
The rationale for WKSIs receiving the most
flexibility is that these firms are so widely followed that investors need less regulatory
protection WKSIs accounted for 95 of the US market capitalization during 2002 to 2009
Even though WKSIs include the largest SEO firms 82 of our sample is composed of mid-cap
(447) and small-cap (375) firms (untabulated)5
Instead of restricting disclosure the SEC
argues that disclosure should be encouraged for these firms (SOR Final Rule p 44731)
SOR relaxes disclosure restrictions for WKSIs in several ways First it removes the
waiting period and changes the pre-filing and post-effective periods by dropping communication
restrictions throughout the entire SEO process (see Fig 1) WKSIs may report in any period
3 Some attorneys post-SOR continued to advise against providing any communications at any time before the SEO
that could be viewed as preparing the market for an offering or that had not undergone due diligence in a prospectus
(Cadwalader Wickersham amp Taft LLP 2005) 4
Alternatively firms with at least $1 billion of nonconvertible securities issuances over the past three years qualify
as WKSIs 5
These figures use the following thresholds for market capitalization $07 billion to $14 billion are small -cap $15
billion to $44 billion are mid-cap $45 billion and above are large-cap
6
factual and forward-looking information that the firm regularly releases6
Second SOR provides
multiple safe harbors for regularly released factual business and forward-looking information7
Third SOR allows WKSIs to freely communicate (ie free write) during any period using an
FWP FWPs include communications in written form and communications via television radio
broadcasts and certain electronic road show slides or handouts (Pena 2007)8
Fourth for
WKSIs shelf registrations are automatically effective without SEC review and the filing period
for shelf registrations is extended from two to three years WKSIs can also update or renew the
shelf registration without SEC review and are no longer responsible for ensuring that investors
receive the final prospectus because filing with the SEC is now viewed as ldquoaccess equals
deliveryrdquo (SOR Final Rule pp 44784-44785) These changes increase a WKSIrsquos ability to
quickly access capital markets when necessary Finally SOR allows firms to omit information
(eg proposed offering amounts for each type of security) from the shelf registration which
effectively shifts the SEO information event from the filing date to the pre-issue date9
As Fig 1
indicates the shelf registration Form S-3ASR allows WKSIs to register to offer securities at any
time without stating detailed information about the SEO or paying registration fees which would
provide some indication of the offering amount10
6 In the SOR Final Rule the SEC clarifies that regularly released forecasts include scheduled forecasts as well as
unscheduled or episodic forecasts The firm could satisfy the regularly released condition as long as the company
has released one prior forecast (SOR Final Rule p 44737) However Cadwalader Wickersham amp Taft LLP a large
international law firm included the following statement in its August 4 2005 memo explaining SOR ldquoThe safe harbors do not establish any minimum time period to satisfy the lsquoregularly releasedrsquo element of the Final Rules While the SEC has stated that one prior release or dissemination could establish a track record if such
communication is the industry norm we believe that commencing a practice for the purpose of meeting this
requirement of the safe harbor will have significant risk until there is more clarification on this pointrdquo (Cadwalader Wickersham amp Taft LLP p 6) 7 Under SORrsquos safe harbor WKSIs are not liable for any forward-looking statement as long as it is accompanied by
certain cautionary language and the WKSI did not have knowledge that the statement was false 8
FWP filings are not required for traditional live road shows in which the SEO firm and underwriters meet potential investors 9
Autore et al (2008) report that 85 of firms filing a shelf registration never issue equity 10
For example firms are not required to state the size of the offering or indicate the specific security that will be issued at the filing date Before SOR the shelf registration included the amount that firms expected to issue over the
7
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Differences in the information environment prior to seasoned equity offerings under
relaxed disclosure regulation
Sarah B Clinton
Department of Accounting and Information Management
College of Business Administration
University of Tennessee
620 Stokely Management Center Knoxville TN 37996
clintonutkedu
Joshua T White
US Securities and Exchange Commission Division of Economic and Risk Analysis
Office of Corporate Finance 100 F Street NE Washington DC 20549
whitejosecgov
Tracie Woidtke
Department of Finance
College of Business Administration
University of Tennessee
425 Stokely Management Center Knoxville TN 37996
twoidtkeutkedu
Telephone (865) 974-1718
Fax (865) 974-1716
April 28 2014
We thank an anonymous referee SP Kothari (editor) Anup Agrawal Cindy Alexander Judy
Beckman Bruce Behn James Chyz Shayne Clinton Donial Dastgir Raquel Fox Joan
Heminway Gerard Hoberg Rebekah McCarty Sarah McVay Alex Nekrasov Marlene Plumlee
Andy Puckett Anup Srivastava Amy Starr Karen Ubell Christian Vossler and seminar
participants at the 2010 American Accounting Association Northeast regional meeting 2011
American Accounting Association annual meeting 2012 Financial Accounting and Reporting
Section midyear meeting Southern Methodist University Texas AampM University University of
Alabama University of Arizona Clemson University Miami University University of
Delaware University of Georgia University of Mississippi University of NebraskandashLincoln
University of Utah US Securities and Exchange Commission (SEC) and the University of
Tennessee Corporate Governance Center for insightful comments and suggestions As a matter
of policy the Securities and Exchange Commission disclaims responsibility for any private
publication or statement by any of its employees Although this paper was reviewed by other
staff members within the Commission the views expressed herein are those of the authors and
do not necessarily reflect those of the Commission the Commissioners or other members of the
Staff
Corresponding author
Differences in the information environment prior to seasoned equity offerings under
relaxed disclosure regulation
ABSTRACT
The Securities Offering Reform (SOR) promulgated by the US Securities and Exchange
Commission (SEC) in 2005 represents a major change in the seasoned equity offering (SEO)
process SOR eases disclosure restrictions and reduces uncertainty regarding disclosures allowed
prior to an SEO The SEC argued that SOR would result in an improved information
environment and benefit capital formation efficiency In contrast critics claimed that SOR would
allow firms to hype their stock before an SEO to the detriment of the information environment It
is also possible that SOR is not effective in generating greater disclosure and affecting the
information environment during equity capital formation because of ongoing concerns such as
the litigious climate in the US This paper is the first to examine differences in disclosure and
the information environment at the time of seasoned equity capital formation under SOR
Overall the results suggest that relaxed disclosure restrictions under SOR are associated with
greater disclosure immediately preceding the SEO issue date and greater disclosure is related to
a richer information environment with capital formation benefits During the month before the
SEO issue date under SOR we find more frequent disclosure of management earnings forecasts
and Form 8-K filings and management earnings forecasts are more accurate In addition we find
disclosure within a week before the SEO issue date under SOR is associated with greater
absolute market-adjusted returns (ie information magnitude) and more positive stock returns
with no reversal afterward (ie capital formation benefits through informative disclosure)
Keywords Securities Offering Reform Securities regulation Seasoned equity offerings
Disclosure Information environment Shelf registration
JEL classification M41 G38 K22
Differences in the information environment prior to seasoned equity offerings under
relaxed disclosure regulation
1 Introduction
Itrsquos a basic precept of American securities law that shareholders should be given the information
they need to evaluate their companies
- Robert Jackson professor of law Columbia University (New York Times April 24 2013)
The information environment plays a critical role in capital market efficiency However
the regulatory framework and the litigious climate within the US often result in firms disclosing
only information that is required (Cadwalader Wickersham amp Taft LLP 2005 Rogers and Van
Buskirk 2009) Until recently disclosure prior to seasoned equity offerings (SEOs) had been
limited by the US Securities and Exchange Commission (SEC) even though investor demand
for information was high Section 5(c) of the Securities Act restricted firm disclosure before an
SEO to ban any attempts to ldquocondition the marketrdquo The SEC promulgated the Securities
Offering Reform (SOR) in 2005 to modernize the SEO process and eliminate restrictions that
were no longer viewed as necessary to protect investors yet prevented the dissemination of
accurate information1
This paper investigates whether SOR is associated with differences in
disclosure and a richer information environment during the equity capital formation period (ie
the period preceding an SEO issue when investors assess the SEO price and decide whether to
invest)
SOR is a unique shift in the regulatory environment for SEOs In addition to removing
restrictions on factual and forward-looking disclosures prior to SEOs SOR provides a safe
harbor for disclosures and allows a novel disclosure venue the free writing prospectus (FWP)
which is any written communication that offers to sell securities that are or will be subject to a
registration statement and can include factual and forward-looking information Critics argued
1 See the Securities Offering Reform Final Rule at httpwwwsecgovrulesfinal33-8591frpdf
1
that eased restrictions under SOR would result in firms providing opportunistic disclosure to
ldquohyperdquo expectations before an SEO (eg Morrissey 2007) However the SEC argued that the
primary benefit of SOR was to provide greater and timelier information flow when investors
decide whether to invest in an SEO (SOR Final Rule p 44792) Despite the increased flexibility
under SOR greater disclosure during equity capital formation may not occur for several reasons
Firms may be concerned that greater disclosure could reveal proprietary information (Ali et al
2014) or establish a disclosure precedent that is difficult to maintain or deviates from a long-
standing policy (Graham et al 2005) Firms may also continue the pre-SOR policy of limiting
disclosure during the capital formation period due to liability concerns such as shareholder
litigation or SEC sanctions for conditioning the market (Cadwalader Wickersham amp Taft LLP
2005) Thus it is not clear that SOR automatically leads to greater disclosure or an improved
information environment prior to SEOs
We analyze a sample of 360 SEOs issued by well-known seasoned issuers (WKSIs) a
new class of issuers established by SOR from 2002 to 2009 To provide the first evidence of
whether disclosure is greater and the information environment is richer under SOR during the
capital formation period we conduct tests using multiple forms of factual and forward-looking
disclosure Specifically we examine management earnings forecasts (MEFs) FWPs 8-K filings
and earnings announcements To capture the capital formation period we focus on the month
(ie 30 days) prior to the SEO issue date The role of disclosure before the issue date is
important as a result of changes in the equity offering process including a shift to shelf
registrations and an increase in the distance between the SEO filing and issue dates under SOR
(Table 1) In addition to our overall analysis we separately examine firms with a policy of
providing MEFs as well as other disclosures (forecasting sample) and firms with a policy of
2
providing only disclosures other than MEFs (non-forecasting sample) to investigate whether
firms behave differently under relaxed disclosure regulation based on prior disclosure practices
We conduct three primary tests First we test whether SOR is associated with more
factual and forward-looking disclosure Second we test whether MEF disclosure is indicative of
a richer information environment under SOR for the forecasting sample We examine MEFs to
see if they are more accurate (ie informative) potentially reducing information asymmetry and
litigation risk or overly optimistic (ie opportunistic) as predicted by critics of SOR Third we
analyze stock prices in two ways to further evaluate the information environment and investigate
capital formation benefits We examine absolute market-adjusted returns to measure the
magnitude of information and signed raw returns to investigate whether disclosure is
opportunistic or informative
Overall the results suggest that relaxed disclosure regulation under SOR is associated
with greater disclosure prior to SEOs which is related to a richer information environment with
capital formation benefits Contrary to criticisms of SOR we do not find evidence that these
disclosures are opportunistic in nature The use of FWPs in the full sample and a greater
proportion of annual MEFs in the forecasting sample under SOR result in a 51 greater
proportion of SEOs preceded by guidance disclosure during the capital formation period The
overall frequency of disclosure is 25 greater and the amount of information provided in the 8-
K immediately before the issue date more than doubles in size Furthermore the number of days
between the last disclosure before an SEO and the issue date is lower under SOR Also MEFs
immediately preceding the SEO issue date are 68 more accurate under SOR (ie have reduced
error and bias) and tend to include more downward than upward revisions In addition the
abnormal magnitude of information reflected in absolute market-adjusted returns is almost twice
3
as large after SOR and the difference is driven by SEOs with disclosure within one week before
the issue date Furthermore stock returns during the capital formation period are greater after
SOR for SEOs with disclosure within one week prior to the issue date with no reversal after the
issue date consistent with enhanced capital formation through informative not opportunistic
disclosure These results support the SECrsquos argument that SOR would result in an improved
information environment and benefit capital formation efficiency
Our paper makes several contributions to the literature First we contribute to the
literature on disclosure and the information environment Specifically our paper is the first to
our knowledge to examine different forms of disclosure including the FWP during the equity
capital formation period Second we contribute to the SEO literature by highlighting changes in
the SEO environment that have important implications for future research For example we
show that the capital formation period is different from the pre-SEO period defined in earlier
studies for the majority of SEOs during our more recent sample period Although earlier studies
frequently designate the filing date as the SEO announcement date (eg Asquith and Mullins
1986 Masulis and Korwar 1986) because few SEOs were shelf-registered studies covering
more recent periods and using the filing date as the SEO announcement date may misinterpret
insignificant returns as less negative reactions to SEOs Less information is conveyed at the
filing date as a result of SOR Finally we contribute to research on regulatory reform While
other studies on relaxed regulation tend to find negative outcomes our results indicate that SOR
increases information flow and smoothes capital formation2
Our paper advances the literature on
disclosure SEOs and regulatory reform and should be of interest to the SEC and firms that
access capital on the secondary market
2 Empirical evidence of negative outcomes is linked to relaxed regulation in disclosure (eg Fernandes et al 2010)
and financial markets (eg Beck et al 2010)
4
The remainder of the paper is organized as follows Section 2 provides a discussion of
SOR and presents our empirical questions Section 3 describes the data and Section 4 outlines
our research design Section 5 presents our results Section 6 concludes the paper
2 Background and empirical questions
21 Before the Securities Offering Reform of 2005
Before SOR the SEO process was commonly divided into three periods pre-filing
waiting and post-effective (see Fig 1 for shelf registration timeline) The pre-filing period began
when a firm contemplated an SEO and ended at the filing date when the registration statement
was filed with the SEC The waiting period began at the filing date and ended when the SEC
declared the registration statement effective During the post-effective period a shelf-registered
SEO firm usually filed a preliminary prospectus (pre-issue date) before the shares were sold
(issue date) The post-effective period was typically short to nonexistent for non-shelf (ie
traditional) registrations because the filing date which was also the pre-issue date and issue date
were very close (Bethel and Krigman 2008)
[Insert Fig 1 near here]
Prior to the reform certain communication was prohibited in the pre-filing and waiting
periods and was limited in the post-effective period The pre-filing and waiting periods were
covered by disclosure restrictions banning any attempt to condition the market Any additional
disclosures in connection with the SEO including forward-looking statements were expressly
forbidden with the exception of a prospectus During the post-effective period factual and
forward-looking disclosure was effectively limited through cumbersome prospectus amendment
requirements and liability concerns (Latham amp Watkins LLP 2005) For example forecasts had
5
to be filed as part of the prospectus Many issuers did not want to include a forecast in a
prospectus because it could be viewed as conditioning the market3
Any disclosure that could be
viewed as an offer (eg an MEF) was limited to statutory prospectus disclosure Furthermore
once the final prospectus was prepared issuers were responsible for ensuring that an investor had
received the final prospectus and the SEC reserved the right to review the final prospectus
which could potentially delay the SEO
22 After the Securities Offering Reform of 2005
SOR became effective December 1 2005 Although SOR affects all issuers it provides
the most significant changes for a newly established class of issuers WKSIs which have a
market capitalization of at least $700 million4
The rationale for WKSIs receiving the most
flexibility is that these firms are so widely followed that investors need less regulatory
protection WKSIs accounted for 95 of the US market capitalization during 2002 to 2009
Even though WKSIs include the largest SEO firms 82 of our sample is composed of mid-cap
(447) and small-cap (375) firms (untabulated)5
Instead of restricting disclosure the SEC
argues that disclosure should be encouraged for these firms (SOR Final Rule p 44731)
SOR relaxes disclosure restrictions for WKSIs in several ways First it removes the
waiting period and changes the pre-filing and post-effective periods by dropping communication
restrictions throughout the entire SEO process (see Fig 1) WKSIs may report in any period
3 Some attorneys post-SOR continued to advise against providing any communications at any time before the SEO
that could be viewed as preparing the market for an offering or that had not undergone due diligence in a prospectus
(Cadwalader Wickersham amp Taft LLP 2005) 4
Alternatively firms with at least $1 billion of nonconvertible securities issuances over the past three years qualify
as WKSIs 5
These figures use the following thresholds for market capitalization $07 billion to $14 billion are small -cap $15
billion to $44 billion are mid-cap $45 billion and above are large-cap
6
factual and forward-looking information that the firm regularly releases6
Second SOR provides
multiple safe harbors for regularly released factual business and forward-looking information7
Third SOR allows WKSIs to freely communicate (ie free write) during any period using an
FWP FWPs include communications in written form and communications via television radio
broadcasts and certain electronic road show slides or handouts (Pena 2007)8
Fourth for
WKSIs shelf registrations are automatically effective without SEC review and the filing period
for shelf registrations is extended from two to three years WKSIs can also update or renew the
shelf registration without SEC review and are no longer responsible for ensuring that investors
receive the final prospectus because filing with the SEC is now viewed as ldquoaccess equals
deliveryrdquo (SOR Final Rule pp 44784-44785) These changes increase a WKSIrsquos ability to
quickly access capital markets when necessary Finally SOR allows firms to omit information
(eg proposed offering amounts for each type of security) from the shelf registration which
effectively shifts the SEO information event from the filing date to the pre-issue date9
As Fig 1
indicates the shelf registration Form S-3ASR allows WKSIs to register to offer securities at any
time without stating detailed information about the SEO or paying registration fees which would
provide some indication of the offering amount10
6 In the SOR Final Rule the SEC clarifies that regularly released forecasts include scheduled forecasts as well as
unscheduled or episodic forecasts The firm could satisfy the regularly released condition as long as the company
has released one prior forecast (SOR Final Rule p 44737) However Cadwalader Wickersham amp Taft LLP a large
international law firm included the following statement in its August 4 2005 memo explaining SOR ldquoThe safe harbors do not establish any minimum time period to satisfy the lsquoregularly releasedrsquo element of the Final Rules While the SEC has stated that one prior release or dissemination could establish a track record if such
communication is the industry norm we believe that commencing a practice for the purpose of meeting this
requirement of the safe harbor will have significant risk until there is more clarification on this pointrdquo (Cadwalader Wickersham amp Taft LLP p 6) 7 Under SORrsquos safe harbor WKSIs are not liable for any forward-looking statement as long as it is accompanied by
certain cautionary language and the WKSI did not have knowledge that the statement was false 8
FWP filings are not required for traditional live road shows in which the SEO firm and underwriters meet potential investors 9
Autore et al (2008) report that 85 of firms filing a shelf registration never issue equity 10
For example firms are not required to state the size of the offering or indicate the specific security that will be issued at the filing date Before SOR the shelf registration included the amount that firms expected to issue over the
7
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Differences in the information environment prior to seasoned equity offerings under
relaxed disclosure regulation
ABSTRACT
The Securities Offering Reform (SOR) promulgated by the US Securities and Exchange
Commission (SEC) in 2005 represents a major change in the seasoned equity offering (SEO)
process SOR eases disclosure restrictions and reduces uncertainty regarding disclosures allowed
prior to an SEO The SEC argued that SOR would result in an improved information
environment and benefit capital formation efficiency In contrast critics claimed that SOR would
allow firms to hype their stock before an SEO to the detriment of the information environment It
is also possible that SOR is not effective in generating greater disclosure and affecting the
information environment during equity capital formation because of ongoing concerns such as
the litigious climate in the US This paper is the first to examine differences in disclosure and
the information environment at the time of seasoned equity capital formation under SOR
Overall the results suggest that relaxed disclosure restrictions under SOR are associated with
greater disclosure immediately preceding the SEO issue date and greater disclosure is related to
a richer information environment with capital formation benefits During the month before the
SEO issue date under SOR we find more frequent disclosure of management earnings forecasts
and Form 8-K filings and management earnings forecasts are more accurate In addition we find
disclosure within a week before the SEO issue date under SOR is associated with greater
absolute market-adjusted returns (ie information magnitude) and more positive stock returns
with no reversal afterward (ie capital formation benefits through informative disclosure)
Keywords Securities Offering Reform Securities regulation Seasoned equity offerings
Disclosure Information environment Shelf registration
JEL classification M41 G38 K22
Differences in the information environment prior to seasoned equity offerings under
relaxed disclosure regulation
1 Introduction
Itrsquos a basic precept of American securities law that shareholders should be given the information
they need to evaluate their companies
- Robert Jackson professor of law Columbia University (New York Times April 24 2013)
The information environment plays a critical role in capital market efficiency However
the regulatory framework and the litigious climate within the US often result in firms disclosing
only information that is required (Cadwalader Wickersham amp Taft LLP 2005 Rogers and Van
Buskirk 2009) Until recently disclosure prior to seasoned equity offerings (SEOs) had been
limited by the US Securities and Exchange Commission (SEC) even though investor demand
for information was high Section 5(c) of the Securities Act restricted firm disclosure before an
SEO to ban any attempts to ldquocondition the marketrdquo The SEC promulgated the Securities
Offering Reform (SOR) in 2005 to modernize the SEO process and eliminate restrictions that
were no longer viewed as necessary to protect investors yet prevented the dissemination of
accurate information1
This paper investigates whether SOR is associated with differences in
disclosure and a richer information environment during the equity capital formation period (ie
the period preceding an SEO issue when investors assess the SEO price and decide whether to
invest)
SOR is a unique shift in the regulatory environment for SEOs In addition to removing
restrictions on factual and forward-looking disclosures prior to SEOs SOR provides a safe
harbor for disclosures and allows a novel disclosure venue the free writing prospectus (FWP)
which is any written communication that offers to sell securities that are or will be subject to a
registration statement and can include factual and forward-looking information Critics argued
1 See the Securities Offering Reform Final Rule at httpwwwsecgovrulesfinal33-8591frpdf
1
that eased restrictions under SOR would result in firms providing opportunistic disclosure to
ldquohyperdquo expectations before an SEO (eg Morrissey 2007) However the SEC argued that the
primary benefit of SOR was to provide greater and timelier information flow when investors
decide whether to invest in an SEO (SOR Final Rule p 44792) Despite the increased flexibility
under SOR greater disclosure during equity capital formation may not occur for several reasons
Firms may be concerned that greater disclosure could reveal proprietary information (Ali et al
2014) or establish a disclosure precedent that is difficult to maintain or deviates from a long-
standing policy (Graham et al 2005) Firms may also continue the pre-SOR policy of limiting
disclosure during the capital formation period due to liability concerns such as shareholder
litigation or SEC sanctions for conditioning the market (Cadwalader Wickersham amp Taft LLP
2005) Thus it is not clear that SOR automatically leads to greater disclosure or an improved
information environment prior to SEOs
We analyze a sample of 360 SEOs issued by well-known seasoned issuers (WKSIs) a
new class of issuers established by SOR from 2002 to 2009 To provide the first evidence of
whether disclosure is greater and the information environment is richer under SOR during the
capital formation period we conduct tests using multiple forms of factual and forward-looking
disclosure Specifically we examine management earnings forecasts (MEFs) FWPs 8-K filings
and earnings announcements To capture the capital formation period we focus on the month
(ie 30 days) prior to the SEO issue date The role of disclosure before the issue date is
important as a result of changes in the equity offering process including a shift to shelf
registrations and an increase in the distance between the SEO filing and issue dates under SOR
(Table 1) In addition to our overall analysis we separately examine firms with a policy of
providing MEFs as well as other disclosures (forecasting sample) and firms with a policy of
2
providing only disclosures other than MEFs (non-forecasting sample) to investigate whether
firms behave differently under relaxed disclosure regulation based on prior disclosure practices
We conduct three primary tests First we test whether SOR is associated with more
factual and forward-looking disclosure Second we test whether MEF disclosure is indicative of
a richer information environment under SOR for the forecasting sample We examine MEFs to
see if they are more accurate (ie informative) potentially reducing information asymmetry and
litigation risk or overly optimistic (ie opportunistic) as predicted by critics of SOR Third we
analyze stock prices in two ways to further evaluate the information environment and investigate
capital formation benefits We examine absolute market-adjusted returns to measure the
magnitude of information and signed raw returns to investigate whether disclosure is
opportunistic or informative
Overall the results suggest that relaxed disclosure regulation under SOR is associated
with greater disclosure prior to SEOs which is related to a richer information environment with
capital formation benefits Contrary to criticisms of SOR we do not find evidence that these
disclosures are opportunistic in nature The use of FWPs in the full sample and a greater
proportion of annual MEFs in the forecasting sample under SOR result in a 51 greater
proportion of SEOs preceded by guidance disclosure during the capital formation period The
overall frequency of disclosure is 25 greater and the amount of information provided in the 8-
K immediately before the issue date more than doubles in size Furthermore the number of days
between the last disclosure before an SEO and the issue date is lower under SOR Also MEFs
immediately preceding the SEO issue date are 68 more accurate under SOR (ie have reduced
error and bias) and tend to include more downward than upward revisions In addition the
abnormal magnitude of information reflected in absolute market-adjusted returns is almost twice
3
as large after SOR and the difference is driven by SEOs with disclosure within one week before
the issue date Furthermore stock returns during the capital formation period are greater after
SOR for SEOs with disclosure within one week prior to the issue date with no reversal after the
issue date consistent with enhanced capital formation through informative not opportunistic
disclosure These results support the SECrsquos argument that SOR would result in an improved
information environment and benefit capital formation efficiency
Our paper makes several contributions to the literature First we contribute to the
literature on disclosure and the information environment Specifically our paper is the first to
our knowledge to examine different forms of disclosure including the FWP during the equity
capital formation period Second we contribute to the SEO literature by highlighting changes in
the SEO environment that have important implications for future research For example we
show that the capital formation period is different from the pre-SEO period defined in earlier
studies for the majority of SEOs during our more recent sample period Although earlier studies
frequently designate the filing date as the SEO announcement date (eg Asquith and Mullins
1986 Masulis and Korwar 1986) because few SEOs were shelf-registered studies covering
more recent periods and using the filing date as the SEO announcement date may misinterpret
insignificant returns as less negative reactions to SEOs Less information is conveyed at the
filing date as a result of SOR Finally we contribute to research on regulatory reform While
other studies on relaxed regulation tend to find negative outcomes our results indicate that SOR
increases information flow and smoothes capital formation2
Our paper advances the literature on
disclosure SEOs and regulatory reform and should be of interest to the SEC and firms that
access capital on the secondary market
2 Empirical evidence of negative outcomes is linked to relaxed regulation in disclosure (eg Fernandes et al 2010)
and financial markets (eg Beck et al 2010)
4
The remainder of the paper is organized as follows Section 2 provides a discussion of
SOR and presents our empirical questions Section 3 describes the data and Section 4 outlines
our research design Section 5 presents our results Section 6 concludes the paper
2 Background and empirical questions
21 Before the Securities Offering Reform of 2005
Before SOR the SEO process was commonly divided into three periods pre-filing
waiting and post-effective (see Fig 1 for shelf registration timeline) The pre-filing period began
when a firm contemplated an SEO and ended at the filing date when the registration statement
was filed with the SEC The waiting period began at the filing date and ended when the SEC
declared the registration statement effective During the post-effective period a shelf-registered
SEO firm usually filed a preliminary prospectus (pre-issue date) before the shares were sold
(issue date) The post-effective period was typically short to nonexistent for non-shelf (ie
traditional) registrations because the filing date which was also the pre-issue date and issue date
were very close (Bethel and Krigman 2008)
[Insert Fig 1 near here]
Prior to the reform certain communication was prohibited in the pre-filing and waiting
periods and was limited in the post-effective period The pre-filing and waiting periods were
covered by disclosure restrictions banning any attempt to condition the market Any additional
disclosures in connection with the SEO including forward-looking statements were expressly
forbidden with the exception of a prospectus During the post-effective period factual and
forward-looking disclosure was effectively limited through cumbersome prospectus amendment
requirements and liability concerns (Latham amp Watkins LLP 2005) For example forecasts had
5
to be filed as part of the prospectus Many issuers did not want to include a forecast in a
prospectus because it could be viewed as conditioning the market3
Any disclosure that could be
viewed as an offer (eg an MEF) was limited to statutory prospectus disclosure Furthermore
once the final prospectus was prepared issuers were responsible for ensuring that an investor had
received the final prospectus and the SEC reserved the right to review the final prospectus
which could potentially delay the SEO
22 After the Securities Offering Reform of 2005
SOR became effective December 1 2005 Although SOR affects all issuers it provides
the most significant changes for a newly established class of issuers WKSIs which have a
market capitalization of at least $700 million4
The rationale for WKSIs receiving the most
flexibility is that these firms are so widely followed that investors need less regulatory
protection WKSIs accounted for 95 of the US market capitalization during 2002 to 2009
Even though WKSIs include the largest SEO firms 82 of our sample is composed of mid-cap
(447) and small-cap (375) firms (untabulated)5
Instead of restricting disclosure the SEC
argues that disclosure should be encouraged for these firms (SOR Final Rule p 44731)
SOR relaxes disclosure restrictions for WKSIs in several ways First it removes the
waiting period and changes the pre-filing and post-effective periods by dropping communication
restrictions throughout the entire SEO process (see Fig 1) WKSIs may report in any period
3 Some attorneys post-SOR continued to advise against providing any communications at any time before the SEO
that could be viewed as preparing the market for an offering or that had not undergone due diligence in a prospectus
(Cadwalader Wickersham amp Taft LLP 2005) 4
Alternatively firms with at least $1 billion of nonconvertible securities issuances over the past three years qualify
as WKSIs 5
These figures use the following thresholds for market capitalization $07 billion to $14 billion are small -cap $15
billion to $44 billion are mid-cap $45 billion and above are large-cap
6
factual and forward-looking information that the firm regularly releases6
Second SOR provides
multiple safe harbors for regularly released factual business and forward-looking information7
Third SOR allows WKSIs to freely communicate (ie free write) during any period using an
FWP FWPs include communications in written form and communications via television radio
broadcasts and certain electronic road show slides or handouts (Pena 2007)8
Fourth for
WKSIs shelf registrations are automatically effective without SEC review and the filing period
for shelf registrations is extended from two to three years WKSIs can also update or renew the
shelf registration without SEC review and are no longer responsible for ensuring that investors
receive the final prospectus because filing with the SEC is now viewed as ldquoaccess equals
deliveryrdquo (SOR Final Rule pp 44784-44785) These changes increase a WKSIrsquos ability to
quickly access capital markets when necessary Finally SOR allows firms to omit information
(eg proposed offering amounts for each type of security) from the shelf registration which
effectively shifts the SEO information event from the filing date to the pre-issue date9
As Fig 1
indicates the shelf registration Form S-3ASR allows WKSIs to register to offer securities at any
time without stating detailed information about the SEO or paying registration fees which would
provide some indication of the offering amount10
6 In the SOR Final Rule the SEC clarifies that regularly released forecasts include scheduled forecasts as well as
unscheduled or episodic forecasts The firm could satisfy the regularly released condition as long as the company
has released one prior forecast (SOR Final Rule p 44737) However Cadwalader Wickersham amp Taft LLP a large
international law firm included the following statement in its August 4 2005 memo explaining SOR ldquoThe safe harbors do not establish any minimum time period to satisfy the lsquoregularly releasedrsquo element of the Final Rules While the SEC has stated that one prior release or dissemination could establish a track record if such
communication is the industry norm we believe that commencing a practice for the purpose of meeting this
requirement of the safe harbor will have significant risk until there is more clarification on this pointrdquo (Cadwalader Wickersham amp Taft LLP p 6) 7 Under SORrsquos safe harbor WKSIs are not liable for any forward-looking statement as long as it is accompanied by
certain cautionary language and the WKSI did not have knowledge that the statement was false 8
FWP filings are not required for traditional live road shows in which the SEO firm and underwriters meet potential investors 9
Autore et al (2008) report that 85 of firms filing a shelf registration never issue equity 10
For example firms are not required to state the size of the offering or indicate the specific security that will be issued at the filing date Before SOR the shelf registration included the amount that firms expected to issue over the
7
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Differences in the information environment prior to seasoned equity offerings under
relaxed disclosure regulation
1 Introduction
Itrsquos a basic precept of American securities law that shareholders should be given the information
they need to evaluate their companies
- Robert Jackson professor of law Columbia University (New York Times April 24 2013)
The information environment plays a critical role in capital market efficiency However
the regulatory framework and the litigious climate within the US often result in firms disclosing
only information that is required (Cadwalader Wickersham amp Taft LLP 2005 Rogers and Van
Buskirk 2009) Until recently disclosure prior to seasoned equity offerings (SEOs) had been
limited by the US Securities and Exchange Commission (SEC) even though investor demand
for information was high Section 5(c) of the Securities Act restricted firm disclosure before an
SEO to ban any attempts to ldquocondition the marketrdquo The SEC promulgated the Securities
Offering Reform (SOR) in 2005 to modernize the SEO process and eliminate restrictions that
were no longer viewed as necessary to protect investors yet prevented the dissemination of
accurate information1
This paper investigates whether SOR is associated with differences in
disclosure and a richer information environment during the equity capital formation period (ie
the period preceding an SEO issue when investors assess the SEO price and decide whether to
invest)
SOR is a unique shift in the regulatory environment for SEOs In addition to removing
restrictions on factual and forward-looking disclosures prior to SEOs SOR provides a safe
harbor for disclosures and allows a novel disclosure venue the free writing prospectus (FWP)
which is any written communication that offers to sell securities that are or will be subject to a
registration statement and can include factual and forward-looking information Critics argued
1 See the Securities Offering Reform Final Rule at httpwwwsecgovrulesfinal33-8591frpdf
1
that eased restrictions under SOR would result in firms providing opportunistic disclosure to
ldquohyperdquo expectations before an SEO (eg Morrissey 2007) However the SEC argued that the
primary benefit of SOR was to provide greater and timelier information flow when investors
decide whether to invest in an SEO (SOR Final Rule p 44792) Despite the increased flexibility
under SOR greater disclosure during equity capital formation may not occur for several reasons
Firms may be concerned that greater disclosure could reveal proprietary information (Ali et al
2014) or establish a disclosure precedent that is difficult to maintain or deviates from a long-
standing policy (Graham et al 2005) Firms may also continue the pre-SOR policy of limiting
disclosure during the capital formation period due to liability concerns such as shareholder
litigation or SEC sanctions for conditioning the market (Cadwalader Wickersham amp Taft LLP
2005) Thus it is not clear that SOR automatically leads to greater disclosure or an improved
information environment prior to SEOs
We analyze a sample of 360 SEOs issued by well-known seasoned issuers (WKSIs) a
new class of issuers established by SOR from 2002 to 2009 To provide the first evidence of
whether disclosure is greater and the information environment is richer under SOR during the
capital formation period we conduct tests using multiple forms of factual and forward-looking
disclosure Specifically we examine management earnings forecasts (MEFs) FWPs 8-K filings
and earnings announcements To capture the capital formation period we focus on the month
(ie 30 days) prior to the SEO issue date The role of disclosure before the issue date is
important as a result of changes in the equity offering process including a shift to shelf
registrations and an increase in the distance between the SEO filing and issue dates under SOR
(Table 1) In addition to our overall analysis we separately examine firms with a policy of
providing MEFs as well as other disclosures (forecasting sample) and firms with a policy of
2
providing only disclosures other than MEFs (non-forecasting sample) to investigate whether
firms behave differently under relaxed disclosure regulation based on prior disclosure practices
We conduct three primary tests First we test whether SOR is associated with more
factual and forward-looking disclosure Second we test whether MEF disclosure is indicative of
a richer information environment under SOR for the forecasting sample We examine MEFs to
see if they are more accurate (ie informative) potentially reducing information asymmetry and
litigation risk or overly optimistic (ie opportunistic) as predicted by critics of SOR Third we
analyze stock prices in two ways to further evaluate the information environment and investigate
capital formation benefits We examine absolute market-adjusted returns to measure the
magnitude of information and signed raw returns to investigate whether disclosure is
opportunistic or informative
Overall the results suggest that relaxed disclosure regulation under SOR is associated
with greater disclosure prior to SEOs which is related to a richer information environment with
capital formation benefits Contrary to criticisms of SOR we do not find evidence that these
disclosures are opportunistic in nature The use of FWPs in the full sample and a greater
proportion of annual MEFs in the forecasting sample under SOR result in a 51 greater
proportion of SEOs preceded by guidance disclosure during the capital formation period The
overall frequency of disclosure is 25 greater and the amount of information provided in the 8-
K immediately before the issue date more than doubles in size Furthermore the number of days
between the last disclosure before an SEO and the issue date is lower under SOR Also MEFs
immediately preceding the SEO issue date are 68 more accurate under SOR (ie have reduced
error and bias) and tend to include more downward than upward revisions In addition the
abnormal magnitude of information reflected in absolute market-adjusted returns is almost twice
3
as large after SOR and the difference is driven by SEOs with disclosure within one week before
the issue date Furthermore stock returns during the capital formation period are greater after
SOR for SEOs with disclosure within one week prior to the issue date with no reversal after the
issue date consistent with enhanced capital formation through informative not opportunistic
disclosure These results support the SECrsquos argument that SOR would result in an improved
information environment and benefit capital formation efficiency
Our paper makes several contributions to the literature First we contribute to the
literature on disclosure and the information environment Specifically our paper is the first to
our knowledge to examine different forms of disclosure including the FWP during the equity
capital formation period Second we contribute to the SEO literature by highlighting changes in
the SEO environment that have important implications for future research For example we
show that the capital formation period is different from the pre-SEO period defined in earlier
studies for the majority of SEOs during our more recent sample period Although earlier studies
frequently designate the filing date as the SEO announcement date (eg Asquith and Mullins
1986 Masulis and Korwar 1986) because few SEOs were shelf-registered studies covering
more recent periods and using the filing date as the SEO announcement date may misinterpret
insignificant returns as less negative reactions to SEOs Less information is conveyed at the
filing date as a result of SOR Finally we contribute to research on regulatory reform While
other studies on relaxed regulation tend to find negative outcomes our results indicate that SOR
increases information flow and smoothes capital formation2
Our paper advances the literature on
disclosure SEOs and regulatory reform and should be of interest to the SEC and firms that
access capital on the secondary market
2 Empirical evidence of negative outcomes is linked to relaxed regulation in disclosure (eg Fernandes et al 2010)
and financial markets (eg Beck et al 2010)
4
The remainder of the paper is organized as follows Section 2 provides a discussion of
SOR and presents our empirical questions Section 3 describes the data and Section 4 outlines
our research design Section 5 presents our results Section 6 concludes the paper
2 Background and empirical questions
21 Before the Securities Offering Reform of 2005
Before SOR the SEO process was commonly divided into three periods pre-filing
waiting and post-effective (see Fig 1 for shelf registration timeline) The pre-filing period began
when a firm contemplated an SEO and ended at the filing date when the registration statement
was filed with the SEC The waiting period began at the filing date and ended when the SEC
declared the registration statement effective During the post-effective period a shelf-registered
SEO firm usually filed a preliminary prospectus (pre-issue date) before the shares were sold
(issue date) The post-effective period was typically short to nonexistent for non-shelf (ie
traditional) registrations because the filing date which was also the pre-issue date and issue date
were very close (Bethel and Krigman 2008)
[Insert Fig 1 near here]
Prior to the reform certain communication was prohibited in the pre-filing and waiting
periods and was limited in the post-effective period The pre-filing and waiting periods were
covered by disclosure restrictions banning any attempt to condition the market Any additional
disclosures in connection with the SEO including forward-looking statements were expressly
forbidden with the exception of a prospectus During the post-effective period factual and
forward-looking disclosure was effectively limited through cumbersome prospectus amendment
requirements and liability concerns (Latham amp Watkins LLP 2005) For example forecasts had
5
to be filed as part of the prospectus Many issuers did not want to include a forecast in a
prospectus because it could be viewed as conditioning the market3
Any disclosure that could be
viewed as an offer (eg an MEF) was limited to statutory prospectus disclosure Furthermore
once the final prospectus was prepared issuers were responsible for ensuring that an investor had
received the final prospectus and the SEC reserved the right to review the final prospectus
which could potentially delay the SEO
22 After the Securities Offering Reform of 2005
SOR became effective December 1 2005 Although SOR affects all issuers it provides
the most significant changes for a newly established class of issuers WKSIs which have a
market capitalization of at least $700 million4
The rationale for WKSIs receiving the most
flexibility is that these firms are so widely followed that investors need less regulatory
protection WKSIs accounted for 95 of the US market capitalization during 2002 to 2009
Even though WKSIs include the largest SEO firms 82 of our sample is composed of mid-cap
(447) and small-cap (375) firms (untabulated)5
Instead of restricting disclosure the SEC
argues that disclosure should be encouraged for these firms (SOR Final Rule p 44731)
SOR relaxes disclosure restrictions for WKSIs in several ways First it removes the
waiting period and changes the pre-filing and post-effective periods by dropping communication
restrictions throughout the entire SEO process (see Fig 1) WKSIs may report in any period
3 Some attorneys post-SOR continued to advise against providing any communications at any time before the SEO
that could be viewed as preparing the market for an offering or that had not undergone due diligence in a prospectus
(Cadwalader Wickersham amp Taft LLP 2005) 4
Alternatively firms with at least $1 billion of nonconvertible securities issuances over the past three years qualify
as WKSIs 5
These figures use the following thresholds for market capitalization $07 billion to $14 billion are small -cap $15
billion to $44 billion are mid-cap $45 billion and above are large-cap
6
factual and forward-looking information that the firm regularly releases6
Second SOR provides
multiple safe harbors for regularly released factual business and forward-looking information7
Third SOR allows WKSIs to freely communicate (ie free write) during any period using an
FWP FWPs include communications in written form and communications via television radio
broadcasts and certain electronic road show slides or handouts (Pena 2007)8
Fourth for
WKSIs shelf registrations are automatically effective without SEC review and the filing period
for shelf registrations is extended from two to three years WKSIs can also update or renew the
shelf registration without SEC review and are no longer responsible for ensuring that investors
receive the final prospectus because filing with the SEC is now viewed as ldquoaccess equals
deliveryrdquo (SOR Final Rule pp 44784-44785) These changes increase a WKSIrsquos ability to
quickly access capital markets when necessary Finally SOR allows firms to omit information
(eg proposed offering amounts for each type of security) from the shelf registration which
effectively shifts the SEO information event from the filing date to the pre-issue date9
As Fig 1
indicates the shelf registration Form S-3ASR allows WKSIs to register to offer securities at any
time without stating detailed information about the SEO or paying registration fees which would
provide some indication of the offering amount10
6 In the SOR Final Rule the SEC clarifies that regularly released forecasts include scheduled forecasts as well as
unscheduled or episodic forecasts The firm could satisfy the regularly released condition as long as the company
has released one prior forecast (SOR Final Rule p 44737) However Cadwalader Wickersham amp Taft LLP a large
international law firm included the following statement in its August 4 2005 memo explaining SOR ldquoThe safe harbors do not establish any minimum time period to satisfy the lsquoregularly releasedrsquo element of the Final Rules While the SEC has stated that one prior release or dissemination could establish a track record if such
communication is the industry norm we believe that commencing a practice for the purpose of meeting this
requirement of the safe harbor will have significant risk until there is more clarification on this pointrdquo (Cadwalader Wickersham amp Taft LLP p 6) 7 Under SORrsquos safe harbor WKSIs are not liable for any forward-looking statement as long as it is accompanied by
certain cautionary language and the WKSI did not have knowledge that the statement was false 8
FWP filings are not required for traditional live road shows in which the SEO firm and underwriters meet potential investors 9
Autore et al (2008) report that 85 of firms filing a shelf registration never issue equity 10
For example firms are not required to state the size of the offering or indicate the specific security that will be issued at the filing date Before SOR the shelf registration included the amount that firms expected to issue over the
7
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
that eased restrictions under SOR would result in firms providing opportunistic disclosure to
ldquohyperdquo expectations before an SEO (eg Morrissey 2007) However the SEC argued that the
primary benefit of SOR was to provide greater and timelier information flow when investors
decide whether to invest in an SEO (SOR Final Rule p 44792) Despite the increased flexibility
under SOR greater disclosure during equity capital formation may not occur for several reasons
Firms may be concerned that greater disclosure could reveal proprietary information (Ali et al
2014) or establish a disclosure precedent that is difficult to maintain or deviates from a long-
standing policy (Graham et al 2005) Firms may also continue the pre-SOR policy of limiting
disclosure during the capital formation period due to liability concerns such as shareholder
litigation or SEC sanctions for conditioning the market (Cadwalader Wickersham amp Taft LLP
2005) Thus it is not clear that SOR automatically leads to greater disclosure or an improved
information environment prior to SEOs
We analyze a sample of 360 SEOs issued by well-known seasoned issuers (WKSIs) a
new class of issuers established by SOR from 2002 to 2009 To provide the first evidence of
whether disclosure is greater and the information environment is richer under SOR during the
capital formation period we conduct tests using multiple forms of factual and forward-looking
disclosure Specifically we examine management earnings forecasts (MEFs) FWPs 8-K filings
and earnings announcements To capture the capital formation period we focus on the month
(ie 30 days) prior to the SEO issue date The role of disclosure before the issue date is
important as a result of changes in the equity offering process including a shift to shelf
registrations and an increase in the distance between the SEO filing and issue dates under SOR
(Table 1) In addition to our overall analysis we separately examine firms with a policy of
providing MEFs as well as other disclosures (forecasting sample) and firms with a policy of
2
providing only disclosures other than MEFs (non-forecasting sample) to investigate whether
firms behave differently under relaxed disclosure regulation based on prior disclosure practices
We conduct three primary tests First we test whether SOR is associated with more
factual and forward-looking disclosure Second we test whether MEF disclosure is indicative of
a richer information environment under SOR for the forecasting sample We examine MEFs to
see if they are more accurate (ie informative) potentially reducing information asymmetry and
litigation risk or overly optimistic (ie opportunistic) as predicted by critics of SOR Third we
analyze stock prices in two ways to further evaluate the information environment and investigate
capital formation benefits We examine absolute market-adjusted returns to measure the
magnitude of information and signed raw returns to investigate whether disclosure is
opportunistic or informative
Overall the results suggest that relaxed disclosure regulation under SOR is associated
with greater disclosure prior to SEOs which is related to a richer information environment with
capital formation benefits Contrary to criticisms of SOR we do not find evidence that these
disclosures are opportunistic in nature The use of FWPs in the full sample and a greater
proportion of annual MEFs in the forecasting sample under SOR result in a 51 greater
proportion of SEOs preceded by guidance disclosure during the capital formation period The
overall frequency of disclosure is 25 greater and the amount of information provided in the 8-
K immediately before the issue date more than doubles in size Furthermore the number of days
between the last disclosure before an SEO and the issue date is lower under SOR Also MEFs
immediately preceding the SEO issue date are 68 more accurate under SOR (ie have reduced
error and bias) and tend to include more downward than upward revisions In addition the
abnormal magnitude of information reflected in absolute market-adjusted returns is almost twice
3
as large after SOR and the difference is driven by SEOs with disclosure within one week before
the issue date Furthermore stock returns during the capital formation period are greater after
SOR for SEOs with disclosure within one week prior to the issue date with no reversal after the
issue date consistent with enhanced capital formation through informative not opportunistic
disclosure These results support the SECrsquos argument that SOR would result in an improved
information environment and benefit capital formation efficiency
Our paper makes several contributions to the literature First we contribute to the
literature on disclosure and the information environment Specifically our paper is the first to
our knowledge to examine different forms of disclosure including the FWP during the equity
capital formation period Second we contribute to the SEO literature by highlighting changes in
the SEO environment that have important implications for future research For example we
show that the capital formation period is different from the pre-SEO period defined in earlier
studies for the majority of SEOs during our more recent sample period Although earlier studies
frequently designate the filing date as the SEO announcement date (eg Asquith and Mullins
1986 Masulis and Korwar 1986) because few SEOs were shelf-registered studies covering
more recent periods and using the filing date as the SEO announcement date may misinterpret
insignificant returns as less negative reactions to SEOs Less information is conveyed at the
filing date as a result of SOR Finally we contribute to research on regulatory reform While
other studies on relaxed regulation tend to find negative outcomes our results indicate that SOR
increases information flow and smoothes capital formation2
Our paper advances the literature on
disclosure SEOs and regulatory reform and should be of interest to the SEC and firms that
access capital on the secondary market
2 Empirical evidence of negative outcomes is linked to relaxed regulation in disclosure (eg Fernandes et al 2010)
and financial markets (eg Beck et al 2010)
4
The remainder of the paper is organized as follows Section 2 provides a discussion of
SOR and presents our empirical questions Section 3 describes the data and Section 4 outlines
our research design Section 5 presents our results Section 6 concludes the paper
2 Background and empirical questions
21 Before the Securities Offering Reform of 2005
Before SOR the SEO process was commonly divided into three periods pre-filing
waiting and post-effective (see Fig 1 for shelf registration timeline) The pre-filing period began
when a firm contemplated an SEO and ended at the filing date when the registration statement
was filed with the SEC The waiting period began at the filing date and ended when the SEC
declared the registration statement effective During the post-effective period a shelf-registered
SEO firm usually filed a preliminary prospectus (pre-issue date) before the shares were sold
(issue date) The post-effective period was typically short to nonexistent for non-shelf (ie
traditional) registrations because the filing date which was also the pre-issue date and issue date
were very close (Bethel and Krigman 2008)
[Insert Fig 1 near here]
Prior to the reform certain communication was prohibited in the pre-filing and waiting
periods and was limited in the post-effective period The pre-filing and waiting periods were
covered by disclosure restrictions banning any attempt to condition the market Any additional
disclosures in connection with the SEO including forward-looking statements were expressly
forbidden with the exception of a prospectus During the post-effective period factual and
forward-looking disclosure was effectively limited through cumbersome prospectus amendment
requirements and liability concerns (Latham amp Watkins LLP 2005) For example forecasts had
5
to be filed as part of the prospectus Many issuers did not want to include a forecast in a
prospectus because it could be viewed as conditioning the market3
Any disclosure that could be
viewed as an offer (eg an MEF) was limited to statutory prospectus disclosure Furthermore
once the final prospectus was prepared issuers were responsible for ensuring that an investor had
received the final prospectus and the SEC reserved the right to review the final prospectus
which could potentially delay the SEO
22 After the Securities Offering Reform of 2005
SOR became effective December 1 2005 Although SOR affects all issuers it provides
the most significant changes for a newly established class of issuers WKSIs which have a
market capitalization of at least $700 million4
The rationale for WKSIs receiving the most
flexibility is that these firms are so widely followed that investors need less regulatory
protection WKSIs accounted for 95 of the US market capitalization during 2002 to 2009
Even though WKSIs include the largest SEO firms 82 of our sample is composed of mid-cap
(447) and small-cap (375) firms (untabulated)5
Instead of restricting disclosure the SEC
argues that disclosure should be encouraged for these firms (SOR Final Rule p 44731)
SOR relaxes disclosure restrictions for WKSIs in several ways First it removes the
waiting period and changes the pre-filing and post-effective periods by dropping communication
restrictions throughout the entire SEO process (see Fig 1) WKSIs may report in any period
3 Some attorneys post-SOR continued to advise against providing any communications at any time before the SEO
that could be viewed as preparing the market for an offering or that had not undergone due diligence in a prospectus
(Cadwalader Wickersham amp Taft LLP 2005) 4
Alternatively firms with at least $1 billion of nonconvertible securities issuances over the past three years qualify
as WKSIs 5
These figures use the following thresholds for market capitalization $07 billion to $14 billion are small -cap $15
billion to $44 billion are mid-cap $45 billion and above are large-cap
6
factual and forward-looking information that the firm regularly releases6
Second SOR provides
multiple safe harbors for regularly released factual business and forward-looking information7
Third SOR allows WKSIs to freely communicate (ie free write) during any period using an
FWP FWPs include communications in written form and communications via television radio
broadcasts and certain electronic road show slides or handouts (Pena 2007)8
Fourth for
WKSIs shelf registrations are automatically effective without SEC review and the filing period
for shelf registrations is extended from two to three years WKSIs can also update or renew the
shelf registration without SEC review and are no longer responsible for ensuring that investors
receive the final prospectus because filing with the SEC is now viewed as ldquoaccess equals
deliveryrdquo (SOR Final Rule pp 44784-44785) These changes increase a WKSIrsquos ability to
quickly access capital markets when necessary Finally SOR allows firms to omit information
(eg proposed offering amounts for each type of security) from the shelf registration which
effectively shifts the SEO information event from the filing date to the pre-issue date9
As Fig 1
indicates the shelf registration Form S-3ASR allows WKSIs to register to offer securities at any
time without stating detailed information about the SEO or paying registration fees which would
provide some indication of the offering amount10
6 In the SOR Final Rule the SEC clarifies that regularly released forecasts include scheduled forecasts as well as
unscheduled or episodic forecasts The firm could satisfy the regularly released condition as long as the company
has released one prior forecast (SOR Final Rule p 44737) However Cadwalader Wickersham amp Taft LLP a large
international law firm included the following statement in its August 4 2005 memo explaining SOR ldquoThe safe harbors do not establish any minimum time period to satisfy the lsquoregularly releasedrsquo element of the Final Rules While the SEC has stated that one prior release or dissemination could establish a track record if such
communication is the industry norm we believe that commencing a practice for the purpose of meeting this
requirement of the safe harbor will have significant risk until there is more clarification on this pointrdquo (Cadwalader Wickersham amp Taft LLP p 6) 7 Under SORrsquos safe harbor WKSIs are not liable for any forward-looking statement as long as it is accompanied by
certain cautionary language and the WKSI did not have knowledge that the statement was false 8
FWP filings are not required for traditional live road shows in which the SEO firm and underwriters meet potential investors 9
Autore et al (2008) report that 85 of firms filing a shelf registration never issue equity 10
For example firms are not required to state the size of the offering or indicate the specific security that will be issued at the filing date Before SOR the shelf registration included the amount that firms expected to issue over the
7
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
providing only disclosures other than MEFs (non-forecasting sample) to investigate whether
firms behave differently under relaxed disclosure regulation based on prior disclosure practices
We conduct three primary tests First we test whether SOR is associated with more
factual and forward-looking disclosure Second we test whether MEF disclosure is indicative of
a richer information environment under SOR for the forecasting sample We examine MEFs to
see if they are more accurate (ie informative) potentially reducing information asymmetry and
litigation risk or overly optimistic (ie opportunistic) as predicted by critics of SOR Third we
analyze stock prices in two ways to further evaluate the information environment and investigate
capital formation benefits We examine absolute market-adjusted returns to measure the
magnitude of information and signed raw returns to investigate whether disclosure is
opportunistic or informative
Overall the results suggest that relaxed disclosure regulation under SOR is associated
with greater disclosure prior to SEOs which is related to a richer information environment with
capital formation benefits Contrary to criticisms of SOR we do not find evidence that these
disclosures are opportunistic in nature The use of FWPs in the full sample and a greater
proportion of annual MEFs in the forecasting sample under SOR result in a 51 greater
proportion of SEOs preceded by guidance disclosure during the capital formation period The
overall frequency of disclosure is 25 greater and the amount of information provided in the 8-
K immediately before the issue date more than doubles in size Furthermore the number of days
between the last disclosure before an SEO and the issue date is lower under SOR Also MEFs
immediately preceding the SEO issue date are 68 more accurate under SOR (ie have reduced
error and bias) and tend to include more downward than upward revisions In addition the
abnormal magnitude of information reflected in absolute market-adjusted returns is almost twice
3
as large after SOR and the difference is driven by SEOs with disclosure within one week before
the issue date Furthermore stock returns during the capital formation period are greater after
SOR for SEOs with disclosure within one week prior to the issue date with no reversal after the
issue date consistent with enhanced capital formation through informative not opportunistic
disclosure These results support the SECrsquos argument that SOR would result in an improved
information environment and benefit capital formation efficiency
Our paper makes several contributions to the literature First we contribute to the
literature on disclosure and the information environment Specifically our paper is the first to
our knowledge to examine different forms of disclosure including the FWP during the equity
capital formation period Second we contribute to the SEO literature by highlighting changes in
the SEO environment that have important implications for future research For example we
show that the capital formation period is different from the pre-SEO period defined in earlier
studies for the majority of SEOs during our more recent sample period Although earlier studies
frequently designate the filing date as the SEO announcement date (eg Asquith and Mullins
1986 Masulis and Korwar 1986) because few SEOs were shelf-registered studies covering
more recent periods and using the filing date as the SEO announcement date may misinterpret
insignificant returns as less negative reactions to SEOs Less information is conveyed at the
filing date as a result of SOR Finally we contribute to research on regulatory reform While
other studies on relaxed regulation tend to find negative outcomes our results indicate that SOR
increases information flow and smoothes capital formation2
Our paper advances the literature on
disclosure SEOs and regulatory reform and should be of interest to the SEC and firms that
access capital on the secondary market
2 Empirical evidence of negative outcomes is linked to relaxed regulation in disclosure (eg Fernandes et al 2010)
and financial markets (eg Beck et al 2010)
4
The remainder of the paper is organized as follows Section 2 provides a discussion of
SOR and presents our empirical questions Section 3 describes the data and Section 4 outlines
our research design Section 5 presents our results Section 6 concludes the paper
2 Background and empirical questions
21 Before the Securities Offering Reform of 2005
Before SOR the SEO process was commonly divided into three periods pre-filing
waiting and post-effective (see Fig 1 for shelf registration timeline) The pre-filing period began
when a firm contemplated an SEO and ended at the filing date when the registration statement
was filed with the SEC The waiting period began at the filing date and ended when the SEC
declared the registration statement effective During the post-effective period a shelf-registered
SEO firm usually filed a preliminary prospectus (pre-issue date) before the shares were sold
(issue date) The post-effective period was typically short to nonexistent for non-shelf (ie
traditional) registrations because the filing date which was also the pre-issue date and issue date
were very close (Bethel and Krigman 2008)
[Insert Fig 1 near here]
Prior to the reform certain communication was prohibited in the pre-filing and waiting
periods and was limited in the post-effective period The pre-filing and waiting periods were
covered by disclosure restrictions banning any attempt to condition the market Any additional
disclosures in connection with the SEO including forward-looking statements were expressly
forbidden with the exception of a prospectus During the post-effective period factual and
forward-looking disclosure was effectively limited through cumbersome prospectus amendment
requirements and liability concerns (Latham amp Watkins LLP 2005) For example forecasts had
5
to be filed as part of the prospectus Many issuers did not want to include a forecast in a
prospectus because it could be viewed as conditioning the market3
Any disclosure that could be
viewed as an offer (eg an MEF) was limited to statutory prospectus disclosure Furthermore
once the final prospectus was prepared issuers were responsible for ensuring that an investor had
received the final prospectus and the SEC reserved the right to review the final prospectus
which could potentially delay the SEO
22 After the Securities Offering Reform of 2005
SOR became effective December 1 2005 Although SOR affects all issuers it provides
the most significant changes for a newly established class of issuers WKSIs which have a
market capitalization of at least $700 million4
The rationale for WKSIs receiving the most
flexibility is that these firms are so widely followed that investors need less regulatory
protection WKSIs accounted for 95 of the US market capitalization during 2002 to 2009
Even though WKSIs include the largest SEO firms 82 of our sample is composed of mid-cap
(447) and small-cap (375) firms (untabulated)5
Instead of restricting disclosure the SEC
argues that disclosure should be encouraged for these firms (SOR Final Rule p 44731)
SOR relaxes disclosure restrictions for WKSIs in several ways First it removes the
waiting period and changes the pre-filing and post-effective periods by dropping communication
restrictions throughout the entire SEO process (see Fig 1) WKSIs may report in any period
3 Some attorneys post-SOR continued to advise against providing any communications at any time before the SEO
that could be viewed as preparing the market for an offering or that had not undergone due diligence in a prospectus
(Cadwalader Wickersham amp Taft LLP 2005) 4
Alternatively firms with at least $1 billion of nonconvertible securities issuances over the past three years qualify
as WKSIs 5
These figures use the following thresholds for market capitalization $07 billion to $14 billion are small -cap $15
billion to $44 billion are mid-cap $45 billion and above are large-cap
6
factual and forward-looking information that the firm regularly releases6
Second SOR provides
multiple safe harbors for regularly released factual business and forward-looking information7
Third SOR allows WKSIs to freely communicate (ie free write) during any period using an
FWP FWPs include communications in written form and communications via television radio
broadcasts and certain electronic road show slides or handouts (Pena 2007)8
Fourth for
WKSIs shelf registrations are automatically effective without SEC review and the filing period
for shelf registrations is extended from two to three years WKSIs can also update or renew the
shelf registration without SEC review and are no longer responsible for ensuring that investors
receive the final prospectus because filing with the SEC is now viewed as ldquoaccess equals
deliveryrdquo (SOR Final Rule pp 44784-44785) These changes increase a WKSIrsquos ability to
quickly access capital markets when necessary Finally SOR allows firms to omit information
(eg proposed offering amounts for each type of security) from the shelf registration which
effectively shifts the SEO information event from the filing date to the pre-issue date9
As Fig 1
indicates the shelf registration Form S-3ASR allows WKSIs to register to offer securities at any
time without stating detailed information about the SEO or paying registration fees which would
provide some indication of the offering amount10
6 In the SOR Final Rule the SEC clarifies that regularly released forecasts include scheduled forecasts as well as
unscheduled or episodic forecasts The firm could satisfy the regularly released condition as long as the company
has released one prior forecast (SOR Final Rule p 44737) However Cadwalader Wickersham amp Taft LLP a large
international law firm included the following statement in its August 4 2005 memo explaining SOR ldquoThe safe harbors do not establish any minimum time period to satisfy the lsquoregularly releasedrsquo element of the Final Rules While the SEC has stated that one prior release or dissemination could establish a track record if such
communication is the industry norm we believe that commencing a practice for the purpose of meeting this
requirement of the safe harbor will have significant risk until there is more clarification on this pointrdquo (Cadwalader Wickersham amp Taft LLP p 6) 7 Under SORrsquos safe harbor WKSIs are not liable for any forward-looking statement as long as it is accompanied by
certain cautionary language and the WKSI did not have knowledge that the statement was false 8
FWP filings are not required for traditional live road shows in which the SEO firm and underwriters meet potential investors 9
Autore et al (2008) report that 85 of firms filing a shelf registration never issue equity 10
For example firms are not required to state the size of the offering or indicate the specific security that will be issued at the filing date Before SOR the shelf registration included the amount that firms expected to issue over the
7
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
as large after SOR and the difference is driven by SEOs with disclosure within one week before
the issue date Furthermore stock returns during the capital formation period are greater after
SOR for SEOs with disclosure within one week prior to the issue date with no reversal after the
issue date consistent with enhanced capital formation through informative not opportunistic
disclosure These results support the SECrsquos argument that SOR would result in an improved
information environment and benefit capital formation efficiency
Our paper makes several contributions to the literature First we contribute to the
literature on disclosure and the information environment Specifically our paper is the first to
our knowledge to examine different forms of disclosure including the FWP during the equity
capital formation period Second we contribute to the SEO literature by highlighting changes in
the SEO environment that have important implications for future research For example we
show that the capital formation period is different from the pre-SEO period defined in earlier
studies for the majority of SEOs during our more recent sample period Although earlier studies
frequently designate the filing date as the SEO announcement date (eg Asquith and Mullins
1986 Masulis and Korwar 1986) because few SEOs were shelf-registered studies covering
more recent periods and using the filing date as the SEO announcement date may misinterpret
insignificant returns as less negative reactions to SEOs Less information is conveyed at the
filing date as a result of SOR Finally we contribute to research on regulatory reform While
other studies on relaxed regulation tend to find negative outcomes our results indicate that SOR
increases information flow and smoothes capital formation2
Our paper advances the literature on
disclosure SEOs and regulatory reform and should be of interest to the SEC and firms that
access capital on the secondary market
2 Empirical evidence of negative outcomes is linked to relaxed regulation in disclosure (eg Fernandes et al 2010)
and financial markets (eg Beck et al 2010)
4
The remainder of the paper is organized as follows Section 2 provides a discussion of
SOR and presents our empirical questions Section 3 describes the data and Section 4 outlines
our research design Section 5 presents our results Section 6 concludes the paper
2 Background and empirical questions
21 Before the Securities Offering Reform of 2005
Before SOR the SEO process was commonly divided into three periods pre-filing
waiting and post-effective (see Fig 1 for shelf registration timeline) The pre-filing period began
when a firm contemplated an SEO and ended at the filing date when the registration statement
was filed with the SEC The waiting period began at the filing date and ended when the SEC
declared the registration statement effective During the post-effective period a shelf-registered
SEO firm usually filed a preliminary prospectus (pre-issue date) before the shares were sold
(issue date) The post-effective period was typically short to nonexistent for non-shelf (ie
traditional) registrations because the filing date which was also the pre-issue date and issue date
were very close (Bethel and Krigman 2008)
[Insert Fig 1 near here]
Prior to the reform certain communication was prohibited in the pre-filing and waiting
periods and was limited in the post-effective period The pre-filing and waiting periods were
covered by disclosure restrictions banning any attempt to condition the market Any additional
disclosures in connection with the SEO including forward-looking statements were expressly
forbidden with the exception of a prospectus During the post-effective period factual and
forward-looking disclosure was effectively limited through cumbersome prospectus amendment
requirements and liability concerns (Latham amp Watkins LLP 2005) For example forecasts had
5
to be filed as part of the prospectus Many issuers did not want to include a forecast in a
prospectus because it could be viewed as conditioning the market3
Any disclosure that could be
viewed as an offer (eg an MEF) was limited to statutory prospectus disclosure Furthermore
once the final prospectus was prepared issuers were responsible for ensuring that an investor had
received the final prospectus and the SEC reserved the right to review the final prospectus
which could potentially delay the SEO
22 After the Securities Offering Reform of 2005
SOR became effective December 1 2005 Although SOR affects all issuers it provides
the most significant changes for a newly established class of issuers WKSIs which have a
market capitalization of at least $700 million4
The rationale for WKSIs receiving the most
flexibility is that these firms are so widely followed that investors need less regulatory
protection WKSIs accounted for 95 of the US market capitalization during 2002 to 2009
Even though WKSIs include the largest SEO firms 82 of our sample is composed of mid-cap
(447) and small-cap (375) firms (untabulated)5
Instead of restricting disclosure the SEC
argues that disclosure should be encouraged for these firms (SOR Final Rule p 44731)
SOR relaxes disclosure restrictions for WKSIs in several ways First it removes the
waiting period and changes the pre-filing and post-effective periods by dropping communication
restrictions throughout the entire SEO process (see Fig 1) WKSIs may report in any period
3 Some attorneys post-SOR continued to advise against providing any communications at any time before the SEO
that could be viewed as preparing the market for an offering or that had not undergone due diligence in a prospectus
(Cadwalader Wickersham amp Taft LLP 2005) 4
Alternatively firms with at least $1 billion of nonconvertible securities issuances over the past three years qualify
as WKSIs 5
These figures use the following thresholds for market capitalization $07 billion to $14 billion are small -cap $15
billion to $44 billion are mid-cap $45 billion and above are large-cap
6
factual and forward-looking information that the firm regularly releases6
Second SOR provides
multiple safe harbors for regularly released factual business and forward-looking information7
Third SOR allows WKSIs to freely communicate (ie free write) during any period using an
FWP FWPs include communications in written form and communications via television radio
broadcasts and certain electronic road show slides or handouts (Pena 2007)8
Fourth for
WKSIs shelf registrations are automatically effective without SEC review and the filing period
for shelf registrations is extended from two to three years WKSIs can also update or renew the
shelf registration without SEC review and are no longer responsible for ensuring that investors
receive the final prospectus because filing with the SEC is now viewed as ldquoaccess equals
deliveryrdquo (SOR Final Rule pp 44784-44785) These changes increase a WKSIrsquos ability to
quickly access capital markets when necessary Finally SOR allows firms to omit information
(eg proposed offering amounts for each type of security) from the shelf registration which
effectively shifts the SEO information event from the filing date to the pre-issue date9
As Fig 1
indicates the shelf registration Form S-3ASR allows WKSIs to register to offer securities at any
time without stating detailed information about the SEO or paying registration fees which would
provide some indication of the offering amount10
6 In the SOR Final Rule the SEC clarifies that regularly released forecasts include scheduled forecasts as well as
unscheduled or episodic forecasts The firm could satisfy the regularly released condition as long as the company
has released one prior forecast (SOR Final Rule p 44737) However Cadwalader Wickersham amp Taft LLP a large
international law firm included the following statement in its August 4 2005 memo explaining SOR ldquoThe safe harbors do not establish any minimum time period to satisfy the lsquoregularly releasedrsquo element of the Final Rules While the SEC has stated that one prior release or dissemination could establish a track record if such
communication is the industry norm we believe that commencing a practice for the purpose of meeting this
requirement of the safe harbor will have significant risk until there is more clarification on this pointrdquo (Cadwalader Wickersham amp Taft LLP p 6) 7 Under SORrsquos safe harbor WKSIs are not liable for any forward-looking statement as long as it is accompanied by
certain cautionary language and the WKSI did not have knowledge that the statement was false 8
FWP filings are not required for traditional live road shows in which the SEO firm and underwriters meet potential investors 9
Autore et al (2008) report that 85 of firms filing a shelf registration never issue equity 10
For example firms are not required to state the size of the offering or indicate the specific security that will be issued at the filing date Before SOR the shelf registration included the amount that firms expected to issue over the
7
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
The remainder of the paper is organized as follows Section 2 provides a discussion of
SOR and presents our empirical questions Section 3 describes the data and Section 4 outlines
our research design Section 5 presents our results Section 6 concludes the paper
2 Background and empirical questions
21 Before the Securities Offering Reform of 2005
Before SOR the SEO process was commonly divided into three periods pre-filing
waiting and post-effective (see Fig 1 for shelf registration timeline) The pre-filing period began
when a firm contemplated an SEO and ended at the filing date when the registration statement
was filed with the SEC The waiting period began at the filing date and ended when the SEC
declared the registration statement effective During the post-effective period a shelf-registered
SEO firm usually filed a preliminary prospectus (pre-issue date) before the shares were sold
(issue date) The post-effective period was typically short to nonexistent for non-shelf (ie
traditional) registrations because the filing date which was also the pre-issue date and issue date
were very close (Bethel and Krigman 2008)
[Insert Fig 1 near here]
Prior to the reform certain communication was prohibited in the pre-filing and waiting
periods and was limited in the post-effective period The pre-filing and waiting periods were
covered by disclosure restrictions banning any attempt to condition the market Any additional
disclosures in connection with the SEO including forward-looking statements were expressly
forbidden with the exception of a prospectus During the post-effective period factual and
forward-looking disclosure was effectively limited through cumbersome prospectus amendment
requirements and liability concerns (Latham amp Watkins LLP 2005) For example forecasts had
5
to be filed as part of the prospectus Many issuers did not want to include a forecast in a
prospectus because it could be viewed as conditioning the market3
Any disclosure that could be
viewed as an offer (eg an MEF) was limited to statutory prospectus disclosure Furthermore
once the final prospectus was prepared issuers were responsible for ensuring that an investor had
received the final prospectus and the SEC reserved the right to review the final prospectus
which could potentially delay the SEO
22 After the Securities Offering Reform of 2005
SOR became effective December 1 2005 Although SOR affects all issuers it provides
the most significant changes for a newly established class of issuers WKSIs which have a
market capitalization of at least $700 million4
The rationale for WKSIs receiving the most
flexibility is that these firms are so widely followed that investors need less regulatory
protection WKSIs accounted for 95 of the US market capitalization during 2002 to 2009
Even though WKSIs include the largest SEO firms 82 of our sample is composed of mid-cap
(447) and small-cap (375) firms (untabulated)5
Instead of restricting disclosure the SEC
argues that disclosure should be encouraged for these firms (SOR Final Rule p 44731)
SOR relaxes disclosure restrictions for WKSIs in several ways First it removes the
waiting period and changes the pre-filing and post-effective periods by dropping communication
restrictions throughout the entire SEO process (see Fig 1) WKSIs may report in any period
3 Some attorneys post-SOR continued to advise against providing any communications at any time before the SEO
that could be viewed as preparing the market for an offering or that had not undergone due diligence in a prospectus
(Cadwalader Wickersham amp Taft LLP 2005) 4
Alternatively firms with at least $1 billion of nonconvertible securities issuances over the past three years qualify
as WKSIs 5
These figures use the following thresholds for market capitalization $07 billion to $14 billion are small -cap $15
billion to $44 billion are mid-cap $45 billion and above are large-cap
6
factual and forward-looking information that the firm regularly releases6
Second SOR provides
multiple safe harbors for regularly released factual business and forward-looking information7
Third SOR allows WKSIs to freely communicate (ie free write) during any period using an
FWP FWPs include communications in written form and communications via television radio
broadcasts and certain electronic road show slides or handouts (Pena 2007)8
Fourth for
WKSIs shelf registrations are automatically effective without SEC review and the filing period
for shelf registrations is extended from two to three years WKSIs can also update or renew the
shelf registration without SEC review and are no longer responsible for ensuring that investors
receive the final prospectus because filing with the SEC is now viewed as ldquoaccess equals
deliveryrdquo (SOR Final Rule pp 44784-44785) These changes increase a WKSIrsquos ability to
quickly access capital markets when necessary Finally SOR allows firms to omit information
(eg proposed offering amounts for each type of security) from the shelf registration which
effectively shifts the SEO information event from the filing date to the pre-issue date9
As Fig 1
indicates the shelf registration Form S-3ASR allows WKSIs to register to offer securities at any
time without stating detailed information about the SEO or paying registration fees which would
provide some indication of the offering amount10
6 In the SOR Final Rule the SEC clarifies that regularly released forecasts include scheduled forecasts as well as
unscheduled or episodic forecasts The firm could satisfy the regularly released condition as long as the company
has released one prior forecast (SOR Final Rule p 44737) However Cadwalader Wickersham amp Taft LLP a large
international law firm included the following statement in its August 4 2005 memo explaining SOR ldquoThe safe harbors do not establish any minimum time period to satisfy the lsquoregularly releasedrsquo element of the Final Rules While the SEC has stated that one prior release or dissemination could establish a track record if such
communication is the industry norm we believe that commencing a practice for the purpose of meeting this
requirement of the safe harbor will have significant risk until there is more clarification on this pointrdquo (Cadwalader Wickersham amp Taft LLP p 6) 7 Under SORrsquos safe harbor WKSIs are not liable for any forward-looking statement as long as it is accompanied by
certain cautionary language and the WKSI did not have knowledge that the statement was false 8
FWP filings are not required for traditional live road shows in which the SEO firm and underwriters meet potential investors 9
Autore et al (2008) report that 85 of firms filing a shelf registration never issue equity 10
For example firms are not required to state the size of the offering or indicate the specific security that will be issued at the filing date Before SOR the shelf registration included the amount that firms expected to issue over the
7
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
to be filed as part of the prospectus Many issuers did not want to include a forecast in a
prospectus because it could be viewed as conditioning the market3
Any disclosure that could be
viewed as an offer (eg an MEF) was limited to statutory prospectus disclosure Furthermore
once the final prospectus was prepared issuers were responsible for ensuring that an investor had
received the final prospectus and the SEC reserved the right to review the final prospectus
which could potentially delay the SEO
22 After the Securities Offering Reform of 2005
SOR became effective December 1 2005 Although SOR affects all issuers it provides
the most significant changes for a newly established class of issuers WKSIs which have a
market capitalization of at least $700 million4
The rationale for WKSIs receiving the most
flexibility is that these firms are so widely followed that investors need less regulatory
protection WKSIs accounted for 95 of the US market capitalization during 2002 to 2009
Even though WKSIs include the largest SEO firms 82 of our sample is composed of mid-cap
(447) and small-cap (375) firms (untabulated)5
Instead of restricting disclosure the SEC
argues that disclosure should be encouraged for these firms (SOR Final Rule p 44731)
SOR relaxes disclosure restrictions for WKSIs in several ways First it removes the
waiting period and changes the pre-filing and post-effective periods by dropping communication
restrictions throughout the entire SEO process (see Fig 1) WKSIs may report in any period
3 Some attorneys post-SOR continued to advise against providing any communications at any time before the SEO
that could be viewed as preparing the market for an offering or that had not undergone due diligence in a prospectus
(Cadwalader Wickersham amp Taft LLP 2005) 4
Alternatively firms with at least $1 billion of nonconvertible securities issuances over the past three years qualify
as WKSIs 5
These figures use the following thresholds for market capitalization $07 billion to $14 billion are small -cap $15
billion to $44 billion are mid-cap $45 billion and above are large-cap
6
factual and forward-looking information that the firm regularly releases6
Second SOR provides
multiple safe harbors for regularly released factual business and forward-looking information7
Third SOR allows WKSIs to freely communicate (ie free write) during any period using an
FWP FWPs include communications in written form and communications via television radio
broadcasts and certain electronic road show slides or handouts (Pena 2007)8
Fourth for
WKSIs shelf registrations are automatically effective without SEC review and the filing period
for shelf registrations is extended from two to three years WKSIs can also update or renew the
shelf registration without SEC review and are no longer responsible for ensuring that investors
receive the final prospectus because filing with the SEC is now viewed as ldquoaccess equals
deliveryrdquo (SOR Final Rule pp 44784-44785) These changes increase a WKSIrsquos ability to
quickly access capital markets when necessary Finally SOR allows firms to omit information
(eg proposed offering amounts for each type of security) from the shelf registration which
effectively shifts the SEO information event from the filing date to the pre-issue date9
As Fig 1
indicates the shelf registration Form S-3ASR allows WKSIs to register to offer securities at any
time without stating detailed information about the SEO or paying registration fees which would
provide some indication of the offering amount10
6 In the SOR Final Rule the SEC clarifies that regularly released forecasts include scheduled forecasts as well as
unscheduled or episodic forecasts The firm could satisfy the regularly released condition as long as the company
has released one prior forecast (SOR Final Rule p 44737) However Cadwalader Wickersham amp Taft LLP a large
international law firm included the following statement in its August 4 2005 memo explaining SOR ldquoThe safe harbors do not establish any minimum time period to satisfy the lsquoregularly releasedrsquo element of the Final Rules While the SEC has stated that one prior release or dissemination could establish a track record if such
communication is the industry norm we believe that commencing a practice for the purpose of meeting this
requirement of the safe harbor will have significant risk until there is more clarification on this pointrdquo (Cadwalader Wickersham amp Taft LLP p 6) 7 Under SORrsquos safe harbor WKSIs are not liable for any forward-looking statement as long as it is accompanied by
certain cautionary language and the WKSI did not have knowledge that the statement was false 8
FWP filings are not required for traditional live road shows in which the SEO firm and underwriters meet potential investors 9
Autore et al (2008) report that 85 of firms filing a shelf registration never issue equity 10
For example firms are not required to state the size of the offering or indicate the specific security that will be issued at the filing date Before SOR the shelf registration included the amount that firms expected to issue over the
7
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
factual and forward-looking information that the firm regularly releases6
Second SOR provides
multiple safe harbors for regularly released factual business and forward-looking information7
Third SOR allows WKSIs to freely communicate (ie free write) during any period using an
FWP FWPs include communications in written form and communications via television radio
broadcasts and certain electronic road show slides or handouts (Pena 2007)8
Fourth for
WKSIs shelf registrations are automatically effective without SEC review and the filing period
for shelf registrations is extended from two to three years WKSIs can also update or renew the
shelf registration without SEC review and are no longer responsible for ensuring that investors
receive the final prospectus because filing with the SEC is now viewed as ldquoaccess equals
deliveryrdquo (SOR Final Rule pp 44784-44785) These changes increase a WKSIrsquos ability to
quickly access capital markets when necessary Finally SOR allows firms to omit information
(eg proposed offering amounts for each type of security) from the shelf registration which
effectively shifts the SEO information event from the filing date to the pre-issue date9
As Fig 1
indicates the shelf registration Form S-3ASR allows WKSIs to register to offer securities at any
time without stating detailed information about the SEO or paying registration fees which would
provide some indication of the offering amount10
6 In the SOR Final Rule the SEC clarifies that regularly released forecasts include scheduled forecasts as well as
unscheduled or episodic forecasts The firm could satisfy the regularly released condition as long as the company
has released one prior forecast (SOR Final Rule p 44737) However Cadwalader Wickersham amp Taft LLP a large
international law firm included the following statement in its August 4 2005 memo explaining SOR ldquoThe safe harbors do not establish any minimum time period to satisfy the lsquoregularly releasedrsquo element of the Final Rules While the SEC has stated that one prior release or dissemination could establish a track record if such
communication is the industry norm we believe that commencing a practice for the purpose of meeting this
requirement of the safe harbor will have significant risk until there is more clarification on this pointrdquo (Cadwalader Wickersham amp Taft LLP p 6) 7 Under SORrsquos safe harbor WKSIs are not liable for any forward-looking statement as long as it is accompanied by
certain cautionary language and the WKSI did not have knowledge that the statement was false 8
FWP filings are not required for traditional live road shows in which the SEO firm and underwriters meet potential investors 9
Autore et al (2008) report that 85 of firms filing a shelf registration never issue equity 10
For example firms are not required to state the size of the offering or indicate the specific security that will be issued at the filing date Before SOR the shelf registration included the amount that firms expected to issue over the
7
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
23 Empirical tests
We investigate whether SOR is associated with greater disclosure and a richer
information environment during the capital formation period through three primary tests First
we investigate whether SOR is associated with more factual and forward-looking disclosure by
comparing different measures of disclosure activity (disclosure occurrence frequency 8-K size
and timing) during the capital formation period before and after SOR was promulgated Prior
research provides several reasons for occurrence frequency and size of disclosure to be greater
during this period deter certain types of litigation (eg Field et al 2005) reduce underpricing
(eg Li and Zhuang 2012) lower equity costs (eg Botosan 1997) lower information
asymmetry (eg Diamond and Verrecchia 1991) continue a prior commitment to disclosure
(eg Graham et al 2005) and condition the market (eg Lang and Lundholm 2000) On the
one hand disclosure should increase with relaxed disclosure restrictions if the benefits of greater
disclosure outweigh the costs under SOR Moreover a firm concerned with omitting material
information may provide greater disclosure (SOR Final Rule pp 44795ndash44796) On the other
hand firms may choose to maintain their existing limited disclosure policy and thus not provide
greater disclosure during the capital formation period under the new rules Firms may be
concerned with inadvertent material misstatements of factual or forward-looking information
leading to litigation Legal firms may also advise firms to disclose no more than what is required
during the capital formation period (Cadwalader Wickersham amp Taft LLP 2005)
We also examine the proximity between disclosure of MEFs FWPs 8-Ks and earnings
announcements and the SEO issue date to investigate whether firms alter the disclosure or SEO
next two years and the registration fees paid on this amount Moreover many pre-SOR firms in our sample list only
one or two types of securities such as common stock and debt securities Many WKSIs in our post-SOR sample list
numerous securities in their shelf registration that have never been issued Thus investors do not know how many
shares of equity a firm may issue until the firm files the preliminary prospectus which (if one is filed) discloses the
SEO size but not price or the final prospectus
8
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
timing or both to affect the distance (ie number of days) under SOR Although firms have
specific deadlines for reporting items on an 8-K evidence by Segal and Segal (2013) shows that
managers engage in strategic timing Furthermore the results in Korajczyk et al (1991) suggest
that firms especially non-forecasting firms strategically time SEOs just after earnings
announcements because they prefer issuing equity when investors are most informed about firm
quality However firms may be less dependent upon earnings announcements (ie greater
distance) since they can provide additional disclosure prior to an SEO under SOR
Second we examine whether more MEFs within the capital formation period is
indicative of a richer information environment under SOR by investigating the accuracy of
MEFs Firms have dueling incentives to provide accurate MEFs prior to an SEO Firms may be
able to issue stock at a higher price following opportunistic disclosure (Lang and Lundholm
2000) However Li and Zhuang (2012) find that highly accurate MEFs reduce the cost of an
SEO MEF accuracy during the month prior to an SEO may also be subject to heightened
scrutiny and litigation risk despite the new safe harbor provisions Improved accuracy would be
consistent with informative voluntary disclosure during the capital formation period while worse
accuracy would be consistent with opportunistic voluntary disclosure as predicted by SORrsquos
critics (eg Morrissey 2007)
Finally we examine two aspects of the information environment reflected in stock prices
First we examine whether the magnitude of information measured as cumulative absolute
market-adjusted returns is greater during the capital formation period All else equal if
disclosure during the capital formation period is associated with greater information absolute
market-adjusted returns should be larger during the month prior to the SEO under SOR If a
larger information magnitude results from greater disclosure immediately preceding the issue
9
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
date then the information magnitude should become larger as the issue date approaches and for
firms providing disclosure within one week prior to the SEO Second we examine whether SOR
is associated with capital formation benefits by cumulating signed raw returns during the capital
formation period To the extent that SOR is associated with increased offering flexibility through
enhanced shelf registration and communication mechanisms SOR should be associated with
capital formation benefits reflected in higher cumulative stock returns during the capital
formation period ie a larger increase in stock price prior to the issue date We examine returns
following the SEO to determine whether any capital formation benefits are associated with
opportunistic or informative disclosure under SOR Greater stock returns prior to the issue with
no reversal afterward would be consistent with more informative disclosure and a richer
information environment during equity capital formation Greater stock returns prior to the issue
with a reversal afterward would be consistent with more opportunistic disclosure and a poorer
information environment during equity capital formation under SOR
24 Capital formation period
We define the capital formation period as the one-month (ie 30-day) period before the
SEO issue date in our tests for several reasons First it includes the SEO period when investors
assess the SEO price and decide to invest Prior literature indicates that firms using disclosure to
influence market expectations time the disclosure immediately preceding a pricing event (eg
Brockman et al 2008 Cheng and Lo 2006 Aboody and Kasznik 2000) Thus this period is
when disclosure is most likely to influence capital formation (eg Korajczyk et al 1991)
Second although most regulations are written with respect to the filing date disclosure was also
effectively limited before the issue date possibly as a result of cumbersome and uncertain
10
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
prospectus amendment requirements or firms viewing the prospectus amendment as a new
registration statement Ruland et al (1990) find that few firms issue equity within one month of a
forecast even though firms are likely to issue equity within three months of a forecast Marquardt
and Wiedman (1998) state that gun-jumping laws discourage forecasts in the period before an
SEO Third although the SEC notes in its comments on the quiet period that federal securities
laws do not define the term it uses a 30-day period prior to the filing date in SOR when
establishing bright-line disclosure restrictions for non-WKSIs
3 Data and sample selection
31 Disclosures analyzed
We study multiple forms of disclosure to investigate factual and forward-looking
information provided during the equity capital formation period We examine two types of
forward-looking disclosure termed ldquoguidancerdquo disclosure in this paper MEFs and forward-
looking FWPs MEFs are one of the most informative voluntary disclosures (Beyer et al 2010)
and we can directly measure accuracy for MEFs An annual (quarterly) MEF is a firm-issued
point or range forecast of annual (quarterly) earnings11
An FWP is a novel disclosure introduced
with SOR12
To broadly examine how SOR affects different types of firms we consider two additional
types of factual and forward-looking disclosure earnings announcements and 8-Ks13
This
11 Similar to Gong et al (2009) we include only point and range forecasts which constitute 94 of MEFs in the
First Call database during our sample period Forecast errors are not as clearly defined for other forms of forecasts
We use the mid-point of range forecasts to examine our accuracy research question and replace point forecasts with
the mid-point when the point forecast is outside the forecast range to correct First Call input errors 12
Although we focus on FWPs with forward-looking information firms may also file an FWP after issuing any
shelf-registered security to update the prospectus We exclude these FWPs because they summarize information
available in the final prospectus and are not forward-looking in nature 13
Examining 8-Ks allows us to look at a fuller set of disclosures 8 -Ks may include press releases of an earnings
announcement or MEF but they also typically contain additional information Our separate results for each of these
11
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
approach allows us to study non-forecasting firms with a policy of providing disclosure besides
MEFs and more thoroughly examine other forms of disclosure by forecasting firms Firms have
less discretion over whether and when to provide earnings announcements and 8-Ks than MEFs
but may provide both factual and forward-looking information in these disclosures Earnings
announcements and 8-Ks are two of the primary mediums for ongoing and periodic disclosure
for publicly listed firms The earnings announcement date is a particularly important event for
resolving information asymmetry prior to an SEO (Korajczyk et al 1991) 8-Ks are intended to
maintain the accuracy and adequacy of information disclosed by firms by providing a continuous
stream of disclosure to investors around significant or material corporate events They constitute
more than half of all SEC filings by public firms and are associated with informative disclosure
that are complementary to periodic reports (Lerman and Livnat 2010 Drake et al 2014) In
addition the size of information in 8-Ks can be directly measured Hanley and Hoberg (2013)
find the size of information releases prior to equity offerings impacts the cost of capital For our
analysis we measure the size of the last 8-K filed prior to the SEO issue date14
To control for
potential bias created by SEC changes to 8-K disclosure requirements in 2004 we also measure
8-K abnormal size as the SEO firm 8-K size less the median non-SEO WKSI 8-K size during the
same period15
types of disclosures indicate that our 8-K results are not driven by the inclusion of earnings announcements and
MEFs in the 8-K sample Few firms provide an MEF and we see no shift in the earnings announcements 14
We expect the 8-K file size to be highly correlated with the amount of information provided Other studies use
word counts to proxy for disclosure size (eg Francis et al 2002) For robustness we conduct a word count on 30
randomly selected 8-Ks around SOR File size is 86 correlated with the number of words in both periods 15
The SEC changed the 8-K disclosure requirements effective August 23 2004 [see Lerman and Livnat (2010) for a
more detailed discussion of these changes] which may impact the size of the 8-K independently from SOR To be
considered a non-SEO WKSI for our additional 8-K size measurements a firm must meet the WKSI criteria for
market capitalization ($700 million) and must not conduct an SEO during the same calendar year as the sa mple firm
12
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
32 Sample selection
We obtain data on SEOs for January 1 2002 through December 31 2009 from the
Securities Data Company (SDC) Global New Issues database MEFs and actual earnings are
from First Call16
8-K and FWP disclosures are from the Wharton Research Data Services
(WRDS) SEC Analytics Suite and from the Electronic Data-Gathering Analysis and Retrieval
(EDGAR) system Earnings announcement dates and financial data are from Compustat Stock
return data are from the Center for Research in Security Prices (CRSP) and institutional
ownership data are from the Thomson Reuters 13F Holdings Database Analyst earnings
forecasts are from the Institutional Brokersrsquo Estimate System (IBES)
We verify offering details through a firmrsquos registration statement preliminary prospectus
and final prospectus all of which are collected from EDGAR We identify WKSIs as SEO firms
that have been publicly traded for at least one year and whose market value of equity for the
calendar year-end before the offering is at least $700 million These criteria correctly identify
over 99 of firms that file as WKSIs after SOR We exclude pure secondary offerings (in which
100 of the shares are sold by individuals in the secondary market and the firm raises no
capital) rights offerings offerings by regulated firms (ie financials and utilities) due to their
different disclosure requirements and offerings by non-US firms Finally we exclude 27 firms
with missing data from CRSP or Compustat and three firms without coverage in First Call17
The
final sample contains 360 SEOs issued by 243 unique firms18
We designate SEOs with issue
dates from January 1 2002 to November 30 2005 as pre-SOR and SEOs with issue dates from
16 Our focus on WKSIs after 2001 reduces the potential First Call bias identified by Chuk et al (2013) They find the
number of MEFs using First Call may be biased downward when examining earlier time periods and firms that are
small and not as widely followed but they find little evidence of bias after the 1990s and for firms that are large or
more widely followed 17
We exclude one observation in which the stock price is less than $100 at the beginning of the fiscal year 18
Although only 25 WKSIs issue equity both before and after SOR we repeat our analysis for this subsample and
find qualitatively similar results
13
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
December 1 2005 through December 31 2009 as post-SOR providing a four-year period on
either side of SOR19
4 Research design
41 Occurrence frequency 8-K size and distance
For our first test examining whether more factual and forward-looking disclosure exists
during the capital formation period under SOR we specify the following equation
Y = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + ε (1)
We estimate Eq (1) for each of our dependent variables (Y) which are measures of
occurrence frequency 8-K size and distance We employ a probit regression to estimate Eq (1)
for occurrence because ordinary least squares (OLS) is an inefficient estimation technique when
the dependent variable is binary rather than continuous (Aldrich and Nelson 1984) We estimate
Eq (1) for frequency size and distance using OLS20
We examine occurrence for SEOs when
the firm provides an annual MEF as well as for those that provide any guidance disclosure We
study frequency before an SEO for the number of guidance disclosures 8-Ks and total
disclosures We consider 8-K size for the last 8-K within the capital formation period Finally
we examine distance which is the number of days between the last disclosure (guidance 8-K
and any of the disclosures we capture) and the SEO issue date Our variable of interest SOR is
equal to one for observations in which the SEO issue date is on or after December 1 2005 We
19 Six SEOs are issued in December 2005 and are included in the post-SOR sample
20 We also estimate Eq (1) for frequency using an OLS regression with the natural log of frequency a Tobit
regression a Poisson regression and a negative binomial regression Our results are robust to each alternative specification
14
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
include control variables found to be associated with disclosure21
Full definitions are presented
in the Appendix With the exception of fiscal period end earnings all measures are calculated at
least 30 days prior to the issue date We use robust standard errors clustered at the firm level
throughout our analysis
42 Management earnings forecast accuracy
Our second test examines the richness of the information environment associated with
differences in MEF disclosure under SOR by measuring MEF accuracy Focusing on MEFs
provided during the month prior to the SEO issue date we estimate accuracy as follows We
measure MEF error as the absolute value of the difference and bias as the signed difference
between the firmrsquos last earnings per share forecast (annual or annualized quarterly) and the actual
earnings per share scaled by the stock price at prior fiscal year-end22
Similar to Ajinkya et al
(2005) we use actual earnings data from First Call to ensure consistency with MEFs In addition
First Callrsquos earnings data reflect street (or core) earnings which are of higher quality than GAAP
earnings for valuation purposes (Gu and Chen 2004)
We test MEF accuracy controlling for other factors that may influence forecast accuracy
using the following equation
Z = α + β1 SOR + β2 Ln(market capitalization) + β3 Institutional ownership
+ β4 Residual analyst following + β5 Analyst dispersion + β6 Litigation risk
+ β7 Book-to-market + β8 Loss + β9 Firm risk + β10 Horizon + ε (2)
21 Prior research (eg Ajinkya et al 2005) often includes major auditor as an additional control variable The
biggest four audit firms audit 96 of our sample As a result we exclude this control variable Because the number
of analysts following a firm is highly correlated with firm size we regress analyst following on Ln(market
capitalization) and retain the residuals as residual analyst following This allows us to control for the component of
analyst following not explained by firm size (Chang et al 2006) We use analyst following as a robustness check
and obtain similar results (not tabulated) 22
We repeat the analysis scaling by earnings per share and obtain similar results (not tabulated)
15
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
We estimate Eq (2) for both of our accuracy dependent variables (Z) error and bias
which are measured for the last MEF for firms providing an MEF during the specified period
prior to the SEO issue date Similar to Eq (1) our variable of interest SOR is equal to one if the
issue date is on or after December 1 2005
43 Information magnitude and stock returns
For our third test we examine stock returns in two ways to provide additional analysis of
the information environment during equity capital formation under SOR First we measure the
magnitude of information released with disclosure by analyzing the absolute value of market-
adjusted returns similar to previous studies (eg Francis et al 2002 Bailey et al 2006)
Specifically we measure magnitude as the cumulative absolute daily returns net of the return on
the CRSP value-weighted index over the [-200] trading day period to proxy for the magnitude of
information reflected in stock prices during the capital formation period with day 0 being the
volume-based corrected issue date23
This 21-trading-day window closely corresponds to the
one-month disclosure window used in our disclosure tests Thus magnitude [-200] captures all
information reflected in stock prices (Tetlock 2010) during the capital formation period
including information conveyed by the disclosures we examine Although Henry and Koski
(2010) find no evidence of manipulative short selling prior to shelf-registered SEOs presumably
because anticipating when these SEOs will occur is difficult we focus on abnormal information
magnitude to control for factors other than capital formation disclosure such as increased trading
23 For analysis of returns during the period surrounding an SEO we employ a volume -based correction methodology
because late day or overnight deals may result in processing the trading volume on the following trading day (ie
the day after the issue date in SDC) Following Corwin (2003) we adjust the issue day later by one day when the
trading volume on the day after the issue date in SDC is more than double the trading volume on the SDC issue date
and when it is more than two times the mean trading volume over the past 250 trading days The issue day is
corrected for 65 of our sample
16
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
volume that may influence magnitude during the period surrounding an SEO We define
abnormal magnitude as magnitude [-200] less magnitude [+1+20]
Second we examine signed stock returns to explore whether SOR is associated with
capital formation benefits and more informative or opportunistic disclosure We cumulate the
signed raw daily stock returns (return) during the [-200] trading day period to measure how
information is incorporated into the stock price during the capital formation period Because
SOR impacts the shelf registration process in addition to permissible disclosure return is likely
to be a lower bound for the net effect of increased disclosure under SOR as the information
content of a shelf registration or SEO announcement shifts from the filing date to the pre-issue
date Thus a negative SEO announcement return is more likely to be included in return under
SOR Similar to abnormal magnitude we measure capital formation benefits through abnormal
return defined as return [-200] less return [+1+20] We further examine the period after the
SEO issue date return [+1+20] to distinguish between informative and opportunistic
disclosure A greater return during the capital formation period followed by a reversal after the
issue date post-SOR would be consistent with opportunistic disclosure under SOR Conversely a
greater return with no reversal afterward would be consistent with informative disclosure24
44 Comparison with other studies
Our research design differs from other studies in important ways First we use the issue
date as the SEO date in response to changes in the SEO environment Earlier studies on pre-SEO
disclosure (eg Marquardt and Wiedman 1998 Lang and Lundholm 2000) refer to the filing
or registration date as the SEO announcement date because their samples are composed of non-
24 To help ensure that our post-SOR period captures any potential reversal due to opportunistic disclosure we extend
the post-SOR period to +60 trading days ie one additional quarter Results are robust
17
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
shelf (traditional) offerings In non-shelf offerings firms convey information about the upcoming
SEO on the registration date and the issue date usually occurs soon after registration (Bethel and
Krigman 2008) In contrast over 80 of our sample issues equity from a shelf registration and
the filing date predates the issue date by 257 days on average (Table 1)25
Second we examine
numerous properties of disclosure during the one-month period prior to the SEO issue date In
addition to MEFs and earnings releases we study other forms of disclosure namely 8-Ks and
FWPs To our knowledge this study is the first to examine these forms of disclosure and the first
to examine the accuracy of MEFs provided during the capital formation period Finally we
analyze the forecasting and non-forecasting samples separately to provide insight into pre-SEO
disclosure activity for firms with and firms without a recent history of forecasting Li and
Zhuang (2012) examine management forecasts primarily in the form of MEFs during the 24-
and 12-month period prior to the filing date for SEOs in a more recent period 1997 to 2006
They find that management forecasts are associated with less underpricing in SEOs by smaller
firms but that no significant relation exists in SEOs by larger firms The insignificant results for
large firms may result from either the potential misidentification of the capital formation period
or the influence of both guidance and non-guidance disclosure on the information environment
Shroff et al (2013) examine disclosure before and after SOR However their focus is more
similar to Lang and Lundholm (2000) and Li and Zhuang (2012) Shroff et al focus on the pre-
filing period include two forms of disclosure (forecasts and press releases) and use a restricted
25 Henry and Koski (2010) and Autore et al (2008) discuss the importance and ch allenges of identifying
announcement dates for SEOs especially for shelf registrations Consistent with these studies we find no significant
market reaction at the filing of shelf registrations The median three-day abnormal pre-issue return is -32 which
is consistent with estimates of the announcement returns for shelf-registered SEOs during previous periods (eg
Autore et al 2008) Because most SEOs are now shelf-registered future researchers will want to focus on the pre-
issue date or issue date and not the filing date as the appropriate information event
18
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
sample with predominantly smaller firms which remain subject to some disclosure restrictions
under SOR
5 Results
51 Descriptive statistics
Table 1 Panel A presents the distribution of SEOs according to the issuing firmrsquos MEF
policy We classify SEOs according to whether the firm has chosen to provide at least one annual
or quarterly point or range MEF or both within one year before the issue date (forecasting
sample) or whether the issuing firm has chosen not to provide an MEF (non-forecasting
26 27sample) Managers issue MEFs within one year of 126 SEOs or 35 of the sample Thirty-
eight percent (33) of SEOs are in the forecasting sample pre-SOR (post-SOR) This is
consistent with firms maintaining an established forecasting policy under SOR Thus greater use
of MEFs during the capital formation period under SOR is likely to be confined to firms with a
previous regular release of MEFs (ie forecasting firms)
[Insert Table 1 near here]
Panel B presents SEO characteristics The percentage of shelf-registered SEOs by WKSIs
increases during our sample period from 75 pre-SOR to 87 post-SOR28
Moreover the
average number of days between the filing and issue dates increases from 196 days pre-SOR to
298 days post-SOR In fact the filing date predates the issue date by at least 90 days for 55 of
the SEOs in our sample indicating that over half our sample would be excluded using
26 We find in our sample that forecasting disclosure policy tends to be static over time This approach therefore
allows for a reasonable time period to identify a firmrsquos existing MEF policy and minimizes any bias resulting from limited coverage in First Call prior to our sample period 27
This proportion is consistent with other studies (eg Beyer et al 2010 Chuk et al 2013) 28
We spoke with a chief financial officer of a Fortune 500 company who indicated that with the ease of shelf registration his company employs a non-shelf offering only when the use of proceeds falls outside ordinary use
19
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
restrictions commonly imposed in studies focusing on the filing date (eg Lang and Lundholm
2000 Shroff et al 2013)29
The average number of days between the filing of a preliminary
prospectus (pre-issue date) and the issue date drops from three to two Thus information
provided at the shelf registration is more likely to be dated and underwriters could have less
time to perform sufficient due diligence when the equity is issued post-SOR (Blackwell et al
1990 Coffee 2005) The average proceeds raised increases from $287 million pre-SOR to $394
million post-SOR However the average SEO size represents just over 12 of the SEO firmrsquos
market value of equity in both periods Panel C presents the average annual number of 8-Ks filed
by WKSIs issuing equity during the fiscal year Non-forecasting firms appear to rely more
heavily upon 8-Ks throughout the year than the forecasting firms but the number of 8-Ks is
greater for both groups over time
Table 2 presents descriptive statistics for our control variables Mean values are presented
for the pre- and post-SOR full forecasting and non-forecasting samples The average market
capitalization analyst following and percentage of firms reporting a loss at fiscal year-end are
similar in both periods Institutional ownership and firm risk are greater in all post-SOR samples
However the proportion of observations in high litigation risk industries is similar in the
forecasting sample and lower in the non-forecasting sample post-SOR The non-forecasting
sample is also made up of smaller firms that are more likely to report a loss have slightly greater
analyst following and have greater analyst forecast dispersion post-SOR
[Insert Table 2 near here]
29 The increased distance between the filing and issue dates is not confined to WKSIs For example in SDC the
filing date predates the issue date during our sample period by at least 90 days for 40 of SEOs by seasoned issuers
firms that are smaller than WKSIs and receive fewer benefits under SOR
20
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
52 Results for occurrence frequency 8-K size and distance
Our first test investigates the various ways firms may use relaxed disclosure restrictions
under SOR to provide more information in the capital formation period Although we do not
expect disclosure to decrease under SOR we use a conservative two-tailed test to present
consistent statistics throughout the paper Table 3 presents univariate results for occurrence
(Panel A) frequency (Panel B) 8-K size (Panel C) and distance (Panel D) Occurrence
frequency and 8-K size are each set equal to zero when no disclosure occurs during the one-
month period prior to the issue date30
No greater occurrence of MEFs is found during the capital
formation period in the post-SOR full sample However annual MEFs are significantly more
likely to occur in the post-SOR forecasting sample relative to the pre-SOR forecasting sample
(451 versus 273) and 127 of the post-SOR forecasting sample provide an FWP For the
post-SOR non-forecasting sample 76 file an FWP during the month before an SEO This may
indicate that firms not typically providing MEFs choose to provide guidance disclosure via an
FWP under SOR as opposed to changing forecasting policy
[Insert Table 3 near here]
Total disclosure frequency is 25 larger post-SOR relative to pre-SOR (308 versus
246) Total disclosure consists of MEFs FWPs 8-Ks and earnings announcements For the
post-SOR forecasting sample higher frequency of total disclosure (358 versus 271) is driven by
the higher frequency of annual MEFs (059 versus 031) and 8-Ks (192 versus 158) and the use
30 Our univariate and multivariate results for frequency and 8-K size are robust to excluding firms without disclosure
during the one-month period prior to the issue date (ie excluding zeroes)
21
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
of FWPs31
For the post-SOR non-forecasting sample total disclosure frequency is greater (283
versus 231) as a result of more frequent 8-Ks (223 versus 182) as well as the use of FWPs
8-K size and 8-K abnormal size are significantly greater post-SOR relative to pre-SOR
(597 versus 254 and 517 versus 221 respectively) The differences are also economically
significant in the forecasting and non-forecasting samples However the statistical significance
weakens in the smaller samples The difference in size is significant in both samples in a one-
tailed test but is only significant in the forecasting sample in a more conservative two-tailed test
Although distance cannot be measured for FWPs pre-SOR they have the shortest
distance during the post-SOR capital formation period averaging 315 days The distance
between the last disclosure of any type during the capital formation period and the issue date is
significantly shorter post-SOR (615 versus 785 days) The shorter distance is primarily found
for guidance (ie MEF or FWP) in the post-SOR forecasting sample and for 8-Ks in the post-
SOR non-forecasting sample The distance during the capital formation period is 485 days
shorter in the post-SOR forecasting sample for last guidance and is 178 days shorter in the post-
SOR non-forecasting sample for last 8-K
Overall the results in Table 3 indicate that SEO firms are providing more information
during the capital formation period under SOR We find that a higher proportion of SEOs in the
post-SOR forecasting sample is preceded by an annual MEF in the month before the issue date
and we find SEOs in both the forecasting and non-forecasting samples are preceded by FWPs
post-SOR Frequency and size of 8-Ks are greater in the post-SOR full sample Pre-SEO
disclosure is also closer before the issue date However the shorter distance results from closer
MEFs in the forecasting sample and closer 8-Ks in the non-forecasting sample post-SOR
31 In untabulated results we verify that the MEF results are not due to general trends in forecasting earnings In
contrast to our results we find the average annual frequency of quarterly MEFs decreases by 27 and the average
annual frequency of annual MEFs remains fairly constant in First Call from the pre- to post-SOR period
22
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
The multivariate results in Panel A of Table 4 are consistent with the univariate results
for occurrence frequency and 8-K size The significantly positive coefficient on SOR for the
forecasting sample in Columns 1ndash2 indicate that the likelihood of providing an annual MEF (p =
009) as well as guidance (p = 006) during the capital formation period is greater under SOR32
In addition Column 3 indicates guidance frequency in the capital formation period is greater
under SOR (p = 004) Overall the sign of the coefficients on the control variables loading
significantly are consistent with previous studies The significantly positive coefficients on SOR
(p = 001) in Columns 4ndash5 for the full sample show that both 8-Ks and total disclosures are more
frequent during the capital formation period under SOR The significantly positive coefficients
on SOR in Columns 6ndash7 for the full sample indicate that 8-K size of the last 8-K preceding the
SEO during the capital formation period is significantly larger under SOR even when
controlling for median size during the same period for WKSIs not issuing equity
[Insert Table 4 near here]
The multivariate results in Panel B provide additional evidence of shorter distance under
SOR for the last disclosure prior to the SEO issue date Columns 1ndash3 present results for last
disclosure restricted to the capital formation period The distance for last guidance is
significantly shorter (-563 days p = 003) and distance for last disclosure of any type is weakly
shorter (-106 days p = 020) under SOR The results in Columns 1ndash3 only partially capture the
relation between distance and SOR because the capital formation period restriction excludes the
last disclosure for a portion of the sample and does not account for the disproportionate
representation of the post-SOR sample Columns 4ndash6 remove the capital formation period
32 In untabulated results for the full sample the coefficient on SOR is not significant in a two-tailed test when
occurrence refers to an annual MEF (p = 017) but becomes significant when occurrence refers to an annual MEF or
FWP (p = 001) Thus greater guidance disclosure in the full sample during the capital formation period under SOR
stems from the use of FWPs and greater use of annual MEFs during the month prior to the SEO issue date
23
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
restriction and present results for last disclosure within one year prior to the issue date When
removing the restriction distance is significantly shorter for last 8-K (-691 days p = 002) and
last disclosure of any type (-372 days p = 004) under SOR Overall the results in Table 4
support the univariate results and indicate that SEO firms are providing more shorter distance
disclosure and thus more information during the month before the SEO issue date 33
53 Results for management earnings forecast accuracy
Table 5 presents univariate results for our second test Panel A presents error and bias
results for the last MEF (annual or annualized quarterly) within the one-month period prior to an
SEO to see if MEFs most likely to influence capital formation are more or less accurate post-
SOR34
The average forecast error is 70 lower under SOR (072 versus 243 p lt 001) In
addition average forecast bias is 83 smaller (133 versus 023 p = 009)
[Insert Table 5 near here]
To further explore the source of greater forecast accuracy Panel B (C) presents the
accuracy results for the first and the last annual (quarterly) MEF along with forecast horizons
Here first annual (quarterly) MEF is the first annual (quarterly) MEF provided during the one
year prior to the SEO issue date and last annual (quarterly) MEF is the last annual (quarterly)
MEF provided during the month before the SEO issue date We also present the difference in the
first and last annual (quarterly) MEF before and after SOR to see if firms are increasing or
decreasing their forecast accuracy as the SEO issue date approaches
33 In untabulated results we re-estimate all of our equations including additional measures that proxy for litigation
risk (Kim and Skinner 2012) to help ensure the SOR variable does not proxy for increased litigation risk Our
results for SOR remain robust when we include these additional measures Results are also robust to the inclusion of
SOR interacted with litigation risk The interaction is insignificant in the re-estimation of Table 4 34
We use annual MEF measures if a firm provides both annual and quarterly MEFs during the one-month period
prior to an SEO
24
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
The results indicate that greater accuracy for the last MEF under SOR is primarily driven
by increasing accuracy from first to last annual MEF and by maintaining as opposed to
decreasing accuracy from first to last quarterly MEF as the issue date approaches For example
error drops significantly (p = 001) from the first (136) to the last (077) annual MEF post-
SOR but not pre-SOR Likewise bias decreases significantly from 071 to 032 post-SOR
but not pre-SOR Thus while accuracy is somewhat greater for the first annual MEF under SOR
the significant improvement occurs with the last annual MEF within one month prior to the SEO
issue date Moreover the last annual MEF averages 175 days prior to the fiscal year-end post-
SOR and is not significantly closer to the fiscal year-end than the last annual MEF pre-SOR
With quarterly MEFs error and bias for the last forecast post-SOR are also significantly lower
than for the last forecast pre-SOR However quarterly MEF accuracy is similar for the first and
the last quarterly MEFs post-SOR Conversely accuracy appears to deteriorate between the first
and the last quarterly MEF pre-SOR For example pre-SOR error (bias) increases from 136
(051) for the first to 234 (106) for the last quarterly MEF The pre-SOR increase in error
(098) is significantly greater than the post-SOR decrease (-009) at the 10 level Horizon
is similar for the pre- and post-SOR last quarterly MEF While greater accuracy does not appear
to be attributed to shorter last MEF horizon we control for horizon in a multivariate setting35
Table 6 presents the multivariate results for MEF accuracy While we focus on accuracy
of last MEFs during the capital formation period in Columns 1 and 3 we also include last MEFs
that are not updated in the capital formation period in Columns 2 and 4 to determine if results are
similar We find SOR is associated with significantly lower error and bias for the last MEF
during the one month and one year prior to the SEO issue date (p = 003 and 003 for error p =
35 In untabulated results we perform the same analysis without restricting the last forecast to the prior month before
the SEO We find similar results for annual MEFs However we find no greater accuracy for quarterly MEFs when
they are not restricted to the prior one-month period
25
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
006 and 001 for bias)36
Overall we find no evidence that managers increase MEF optimism
under SOR In fact the results suggest that managers provide less optimistic and more accurate
MEFs before an SEO post-SOR despite the temptation of temporarily influencing stock prices
through rosy forecasts In untabulated results we find that MEFs provided during the month
before the issue date post-SOR have a greater proportion of downward revisions which gives
additional evidence against opportunistic MEFs Falling profits during the 2008 financial crisis
do not appear to explain the downward revisions Less than 6 of the post-SOR subsample with
downward revisions ie SEOs with last MEFs lower than first MEFs have issue dates in 2008
Further we find that post-SOR downward revisers tend to raise larger SEO proceeds and have
greater firm risk both of which are associated with greater litigation risk These results suggest
instead that firms may be correcting expectations downward prior to an SEO to mitigate
litigation risk37
Although managers may delay releasing bad news on average (Kothari et al
2009) fear of litigation may prompt revising MEFs downward prior to an SEO (Skinner 1994)
[Insert Table 6 near here]
54 Results for information magnitude and stock returns
For our third test Fig 2 plots the cumulative magnitude (ie absolute market-adjusted
returns) beginning 20 trading days before and ending on the SEO issue date which approximates
the capital formation period for the pre- and post-SOR periods Consistent with greater
information reflected in stock prices as the issue date approaches under SOR the cumulative
36 In untabulated results we interact SOR with litigation risk All results are robust and the interaction is significant
only when estimating bias for MEFs provided during the prior month The coefficients on SOR and litigation risk
remain negative and significant while the interaction is significantly positive indicating SOR and litigation risk are
both associated with lower bias but the lower bias after SOR is outside litigation risk industries 37
Previous studies find increased earnings management prior to SEOs (eg Cohen and Zarowin 2010) We do not
believe earnings management is responsible for greater accuracy under SOR for two reasons First street earnings
provide more accurate operating earnings for valuation Second we find a greater incidence of downward revisions
instead of upward revisions prior to the SEO
26
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
magnitude post-SOR plot lies above the pre-SOR plot and the gap widens as the issue date
approaches The univariate results in Table 7 Panel A indicate that the post-SOR magnitude [-
200] of 5520 is 47 greater at issue than the pre-SOR value of 3756 (p lt 001) Moreover
the post-SOR abnormal magnitude of 1602 is 87 greater than the pre-SOR value of 855
indicating that the magnitude of information during the capital formation period is significantly
greater under SOR even when adjusting for factors other than capital formation disclosure that
may influence magnitude during the period surrounding an SEO
[Insert Fig 2 and Table 7 near here]
Panel B compares magnitude for the pre-SOR full sample with post-SOR subsamples
based on whether the SEO is preceded by any type of disclosure or an 8-K within one week
defined as five days to see if the increasing magnitude of information as the issue date
approaches post-SOR in Fig 2 is related to disclosure immediately preceding the issue date
Taken together the results for magnitude [-200] and abnormal magnitude provide additional
support that greater information (ie absolute market-adjusted returns) during the capital
formation period under SOR is associated with disclosure Magnitude [-200] is significantly
greater post-SOR regardless of whether SEOs have short distance disclosure (ie disclosure
within one week before the issue date) However abnormal magnitude is significantly greater
only when SEOs have short distance disclosure Average abnormal magnitude is 2008
(2102) when an SEO is immediately preceded by any form of disclosure (an 8-K) post-SOR
compared with 855 for the pre-SOR sample38
38 We analyze the content distribution of the various items reported on the last 8 -K filed within a week prior to the
issue date (short distance 8-K) post-SOR to provide a contextual analysis of these communications (untabulated)
We find over 80 of these 8-Ks include Financial Statements and Exhibits (Item 901) and about 20 include one
or more of the following Regulation FD Disclosure (Item 701) Results of Operations and Financial Condition
(Item 202) and Entry into a Material Definitive Agreement (Item 101) the last of which includes underwriting
agreements related to the SEO More than half include Other Events (Item 801) The SEC encourages firms to
disclose information on Item 801 that is of importance to security holders but disclosure of these items is not
27
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Fig 3 provides additional evidence that greater magnitude post-SOR is related to 8-Ks
during the capital formation period The figure plots three-day [-1+1] magnitude aggregated for
all 8-K filing dates during the capital formation period pre- and post-SOR Similar to Fig 2 it
shows that 8-Ks are associated with greater information magnitude being reflected in stock
prices post-SOR with the difference increasing as the SEO issue date approaches
[Insert Fig 3 near here]
Results for return are presented in Table 839
Overall abnormal return (Panel A) is more
than five times greater under SOR (472 versus 085) The difference is significantly greater
in a one-tailed (two-tailed) test for the full (non-forecasting) sample Greater abnormal return
results from greater return [-200] 503 versus 245 combined with a lower positive drift
after the issue return [+1+20] of 031 versus 160 Similar to the results for magnitude
greater abnormal return is driven by SEOs preceded by any form of disclosure or 8-Ks within
one week Abnormal return (Panel B) is significantly greater in a two-tailed test for the post-
SOR subsamples with any disclosure or an 8-K immediately preceding the SEO (748 and
793 versus 085 pre-SOR) while the differences are insignificant for the post-SOR
subsamples with no disclosure or no 8-K within one week (188 and 188 versus 085 pre-
SOR) Furthermore greater abnormal return in the post-SOR subsamples is primarily driven by
greater positive stock returns during the capital formation period (return [-200] of 836 and
936 versus 245 pre-SOR) No evidence exists of reversal after the issue date in these
subsamples (return [+1+20] is 088 and 143) Higher stock returns during the capital
mandatory (Lerman and Livnat 2010 Segal and Segal 2013) We find that Other Events on these 8-Ks include
information such as press releases MEFs and other forward-looking information These results suggest firms are
providing a variety of information that appears to be important to equity investors during the capital formation
period under SOR 39
We use raw returns because we are interested in the net effects for the firms However we find similar results
when we use cumulative market-adjusted returns In untabulated results we also control for size and book-to-market
in a multivariate setting and continue to find significant results for SOR interacted with short distance disclosure
28
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
formation period with no reversal afterward is consistent with greater capital formation benefits
through more informative disclosure and a richer information environment under SOR
[Insert Table 8 near here]
6 Conclusion
This paper is the first to examine differences in disclosure and the information
environment at the time of seasoned equity capital formation under the Securities Offering
Reform which relaxes disclosure restrictions prior to an SEO We examine whether SOR is
associated with more factual and forward-looking disclosure by examining 8-K file size and the
occurrence frequency and timing of multiple forms of disclosure for a sample of WKSIs who
are the most economically significant issuers and receive the greatest benefits under SOR We
also investigate whether SOR is associated with a more informative or opportunistic information
environment by analyzing management earnings forecast accuracy as well as cumulative
absolute market-adjusted returns and cumulative signed raw returns
Our research design differs from other studies in important ways First we use the issue
date rather than the filing date as the SEO date in response to changes in the registration process
under SOR Second we examine multiple forms of disclosure during the one-month period prior
to the SEO issue date ie the capital formation period In addition to MEFs and earnings
releases we analyze 8-Ks and a new form of disclosure introduced with SOR free writing
prospectuses Finally our sample includes non-forecasting firms in addition to forecasting firms
which allows us to provide insight into capital formation period disclosure by both firms with
and firms without a history of forecasting
29
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
We find greater forward-looking and factual disclosure during the equity capital
formation period under SOR However despite the predictions by critics of SOR we find no
evidence that relaxed disclosure restrictions under SOR are associated with opportunistic
disclosure Within the post-SOR forecasting sample MEFs are more accurate and greater
disclosure and closer proximity are strongest for guidance disclosure In contrast greater
disclosure and closer proximity in the post-SOR non-forecasting sample are primarily driven by
8-K disclosure though some non-forecasting firms do provide guidance disclosure through the
newly permitted FWPs We also find that greater absolute market-adjusted returns under SOR
are highest in the non-forecasting sample and for SEOs with an 8-K filed within a week of the
issue date Moreover significantly more positive stock returns are evident during the capital
formation period with no reversal afterward for SEOs with an 8-K filed within a week of the
issue date This is of particular interest given that the majority of SEO firms continue to provide
no MEFs under SOR One caveat to our results is that the power of certain tests may be limited
by sample size during our sample period As the period since SOR lengthens two fruitful areas
for future research would be to investigate whether disclosure under SOR changes over time and
whether SOR affects a firmrsquos ability to raise capital in different economic cycles
Overall the results in this paper suggest that relaxed disclosure restrictions and reduced
regulatory uncertainty under SOR are associated with greater disclosure immediately preceding
the SEO issue date when investors assess the SEO price and decide whether to invest
Disclosure during this time especially 8-K disclosure is related to a richer information
environment with capital formation benefits Thus our paper contributes to the literature on
disclosure SEOs and regulation
30
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Appendix Variable definitions and other key terms
Variable Definition
Abnormal magnitude
Abnormal return
Analyst dispersion
Analyst following
Annual MEF
Bias
Book-to-market
Closest disclosure
Distance
Earnings announcement
Error
Firm risk
First annual MEF
First quarterly MEF
Frequency
Forecasting
FWP
Guidance
Horizon
Institutional ownership
Last annual MEF
Last MEF
The magnitude over trading days [-200] less the magnitude over trading days [+1+20] The return over trading days [-200] less the return over trading days [+1+20] The standard deviation of analystsrsquo annual earnings estimates divided by the absolute value of the median analyst earnings forecast Analyst dispersion is calculated using the last reported analyst estimates in Institutional Brokersrsquo Estimate System (IBES) for the month-end at least 30 days prior to the seasoned equity offering (SEO) issue date The number of analysts providing annual forecasts at the month ending at least 30 days prior to the SEO issue date A firm-issued point or median of the range forecast of annual earnings The firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The book value of common equity divided by the market value of common equity at the fiscal year-end prior to the SEO issue date Book-to-market is set to the minimum sample value for two observations with missing values The number of days between the last disclosure of the following type prior to the SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement The number of days between the specified disclosure and the SEO issue date The periodic announcement of quarterly or annual earnings during the 30 days prior to the SEO issue date The absolute value of the firmrsquos annual MEF or quarterly MEF minus the actual earnings divided by the stock price at the beginning of the fiscal period annualized The standard deviation of the firms daily returns over the period [-250-20] trading days to the SEO issue date The annual MEF issued farthest from the SEO within 365 days of the SEO issue date The quarterly MEF issued farthest from the SEO within 365 days of the SEO issue date The number of the specified disclosures during the month prior to the SEO issue date An indicator variable equal to one if the firm provides an annual MEF or quarterly MEF during the 365 days prior to the SEO issue date A free writing prospectus is an SEC filing of communication FWPs can include any written communication in printed form and other communication via television radio broadcasts and certain electronic road shows (Pena 2007) A forward-looking disclosure in the form of an MEF or FWP The number of days between the annual MEF or quarterly MEF and the fiscal period end annualized The percentage of the firmrsquos outstanding shares held by institutions at the quarter-end prior to the SEO issue date The annual MEF issued closest to the SEO within 30 days of the SEO issue date The last quarterly MEF or annual MEF within 30 days of the SEO issue date
31
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Last quarterly MEF The quarterly MEF issued closest to the SEO within 30 days of the SEO issue date
Litigation risk An indicator variable equal to one if the SEO firm operates in the following high litigation risk industries (standard
industrial classification codes) biotechnology (2833ndash2836) research and development services (8731ndash8734)
programming (7371ndash7379) computers (3570ndash3577) electronics (3600ndash3674) or retailing (5200ndash5961)
Ln(market capitalization) The natural log of the market value of common equity at fiscal year-end prior to the SEO issue date
Loss An indicator variable equal to one if the firm reports a loss in the current fiscal year
Magnitude The absolute value of the daily return net of the return on the CRSP value-weighted index cumulated over the
specified trading days [-200] and [+0+20] with day 0 being the volume-based corrected issue date (Corwin 2003)
Market capitalization The market value of common equity at the fiscal year-end prior to the SEO issue date in billions of dollars
MEF A management earnings forecast a firm-issued point or median of range forecast of annual or quarterly earnings
Non-forecasting An indicator variable equal to one if the firm does not provide an annual MEF or quarterly MEF during the 365 days
Non-SEO 8-K size
prior to the SEO issue date
The median 8-K size for non-SEO well-known seasoned issuers (WKSIs) during the same period Non-SEO WKSIs
meet the WKSI market capitalization criteria but do not conduct an SEO during the same calendar year
Occurrence An indicator variable equal to one if the firm issues the specified voluntary disclosure (annual MEF quarterly MEF
or FWP) within 30 days prior to the SEO issue date
Pre-issue days The number of days between the filing of a preliminary prospectus (pre-issue date) and the SEO issue date
Quarterly MEF A firm-issued point or median of range forecast of quarterly earnings
Registration days The number of days between the filing of a registration statement and the SEO issue date
Residual analyst following The residual value after regressing analyst following on Ln(market capitalization)
Return
SEO proceeds
SEO size
Shelf
Short distance 8-K
Short distance disclosure
SOR
Total disclosure
The raw daily return cumulated over the specified trading days [-200] and [+0+20] with day 0 being the volume-
based corrected issue date (Corwin 2003)
The gross proceeds from the seasoned equity offering before underwriting
The SEO proceeds scaled by market capitalization at the fiscal year-end prior to the SEO issue date
An indicator variable equal to one if the SEO is offered from a shelf registration
An indicator variable equal to one if the firm provides an 8-K during the five days prior to the SEO issue date
An indicator variable equal to one if the firm provides any of the following disclosures during the five days prior to the
SEO issue date annual MEF quarterly MEF FWP 8-K or earnings announcement
An indicator variable equal to one if the SEO issue date is on or after December 1 2005 when SOR became effective
The aggregate number of annual MEFs quarterly MEFs FWPs 8-Ks and earnings announcements within 30 days of
the SEO issue date
8-K abnormal size The SEO firmrsquos 8-K size less the median non-SEO 8-K size
8-K size The size in megabytes of the last 8-K during the 30 days prior to the SEO issue date otherwise set to zero
32
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
References
Aboody D Kasznik R 2000 CEO stock option awards and the timing of voluntary corporate
disclosures J of Account and Econ 29 (1) 73ndash100
Ajinkya B Bhojraj S Sengupta P 2005 The association between outside directors institutional
investors and the properties of management earnings forecasts J of Account Res 43 (3) 343ndash 376
Aldrich JH Nelson FD 1984 Linear Probability Logit and Probit Models SAGE Publications
Beverly Hills CA
Ali A Klasa S Yeung E 2014 Industry concentration and corporate disclosure policy Working
paper The University of Texas at Dallas
Asquith P Mullins DW 1986 Equity issues and offering dilution J of Financ Econ 15 (1) 61ndash89
Autore DM Kumar R Shome DK 2008 The revival of shelf-registered corporate equity offerings
J of Corp Financ 14 (1) 32ndash50
Bailey W Karolyi GA Salva C 2006 The economic consequences of increased disclosure Evidence
from international cross-listings J of Financ Econ 81 (1) 175ndash213
Beck T Levine R Levkov A 2010 Big bad banks The winners and losers from bank deregulation in
the United States J of Financ 65 (5) 1637ndash1667
Bethel JE Krigman L 2008 Managing the costs of issuing common equity The role of registration
choice Q J of Financ and Account 47 (4) 57ndash85
Beyer A Cohen DA Lys TZ Walther BR 2010 The financial reporting environment Review of
the recent literature J of Account and Econ 50 (2ndash3) 179ndash234
Blackwell DW Marr MW Spivey MF 1990 Shelf registration and the reduced due diligence
argument Implications of the underwriter certification and the implicit insurance hypotheses J
of Financ and Quant Anal 25 (2) 245ndash259
Botosan CA 1997 Disclosure level and the cost of equity capital Account Rev 72 (3) 323ndash349
Brockman P Khurana IK Martin X 2008 Voluntary disclosures around share repurchases J of
Financ Econ 89 (1) 175ndash191
Cadwalader Wickersham amp Taft LLP 2005 New SEC rules for Securities Offering Reform August 4
httpwwwcadwaladercomassetsclient_friend080405SECRulesforSecuritiesOffering
Reformpdf [Accessed 16 Dec 2011]
Chang X Dasgupta S Hillary G 2006 Analyst coverage and financing decisions J of Financ 61 (6)
2009ndash2048
Cheng Q Lo K 2006 Insider trading and voluntary disclosures J of Account Res 44 (5) 815ndash848
33
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Chuk E Matsumoto D Miller G 2013 Assessing methods of identifying management forecasts CIG
vs researcher collected J of Account and Econ 55 (1) 23ndash42
Coffee JC 2005 A Section 11 safe harbor NY Law J 234 5ndash9
Cohen DA Zarowin P 2010 Accrual-based and real earnings management activities around seasoned
equity offerings J of Account and Econ 50 (1) 2ndash19
Corwin SA 2003 The determinants of underpricing for seasoned equity offers J of Financ 58 (5)
2249ndash2279
Diamond DW Verrecchia RE 1991 Disclosure liquidity and the cost of capital J of Financ 46
(4) 1325ndash1359
Drake MS Roulstone DT Thornock JR 2014 The determinants and consequences of information
acquisition via EDGAR Contemp Account Res forthcoming
Fernandes N Lei U Miller DP 2010 Escape from New York The market impact of loosening
disclosure requirements J of Financ Econ 95 (2) 129ndash147
Field L Lowry M Shu S 2005 Does disclosure deter or trigger litigation J of Account and Econ
39 (3) 487ndash507
Francis J Schipper K Vincent L 2002 Earnings announcements and competing information J of
Account and Econ 33 (3) 313ndash342
Gong G Li LY Xie H 2009 The association between management earnings forecasts and accruals
Account Rev 84 (2) 497ndash530
Graham JR Harvey CR Rajgopal S 2005 The economic implications of corporate financial
reporting J of Account and Econ 40 (1ndash3) 3ndash73
Gu Z Chen T 2004 Analystsrsquo treatment of nonrecurring items in street earnings J of Account and Econ 38 129ndash170
Hanley KW Hoberg G 2013 The information content of IPO prospectuses Rev of Financ Stud 23
(7) 2821ndash2864
Henry TR Koski JL 2010 Short selling around seasoned equity offerings Rev of Financ Stud 23
(12) 4389ndash4418
Kim I Skinner DJ 2012 Measuring securities litigation risk J of Account and Econ 53 (1ndash2) 290ndash 310
Korajczyk RA Lucas DJ McDonald RL 1991 The effect of information releases on the pricing
and timing of equity issues Rev of Financ Stud 4 (4) 685ndash708
Kothari SP Shu S Wysocki PD 2009 Do managers withhold bad news J of Account Res 47 (1)
241ndash276
34
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Lang MH Lundholm RJ 2000 Voluntary disclosure and equity offerings Reducing information
asymmetry or hyping the stock Contemp Account Res 17 (4) 623ndash662
Latham amp Watkins LLP 2005 Christmas in JulyndashThe SEC improves the securities offering process
August httpwwwlwcomuploadpubcontent_pdfpub1360_1pdf [Accessed 25 Jun 2013]
Lerman A Livnat J 2010 The new Form 8-K disclosures Rev of Account Stud 15 (4) 752ndash778
Li OZ Zhuang Z 2012 Management guidance and the underpricing of seasoned equity offerings
Contemp Account Res 29 (3) 710ndash737
Marquardt CA Wiedman CI 1998 Voluntary disclosure information asymmetry and insider selling
through secondary equity offerings Contemp Account Res 15 (4) 505ndash537
Masulis RW Korwar AN 1986 Seasoned equity offerings An empirical investigation J of Financ
Econ 15 (1) 91ndash118
Morrissey JF 2007 Rhetoric and reality Investor protection and the securities regulation reform of
2005 Cathol Univ Law Rev 56 (2) 561ndash607
Pena A 2007 The free writing prospectus A six-question approach for issuers Secur Regul Law J 35
(1) 36ndash70
Rogers JL Van Buskirk A 2009 Shareholder litigation and changes in disclosure behavior J of
Account and Econ 47 (1ndash2) 136ndash156
Ruland W Tung S George NE 1990 Factors associated with the disclosure of managersrsquo forecasts
Account Rev 65 (3) 710ndash721
Segal B Segal D 2013 The opportunistic reporting of material events and the apparent misconception
of investorsrsquo reaction Working paper INSEAD
Shroff NO Sun AX White HD Zhang W 2013 Voluntary disclosure and information
asymmetry Evidence from the 2005 Securities Offering Reform J of Account Res 51 (5)
1261ndash1345
Skinner DJ 1994 Why firms voluntarily disclose bad news J of Account Res 32 (1) 38ndash60
Tetlock PC 2010 Does public financial news resolve asymmetric information Rev of Financ Stud
23 (9) 3520ndash3557
35
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Fig 1 Timeline of shelf registration and issue before and after the Securities Offering
Reform (SOR)
Pre-SOR
Post-SOR
Denotes changes from pre-SOR
36
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
100
90
Cu
mu
lati
ve
ma
gn
itu
de
(per
cen
t)
80
70
60
50
40
30
20
10
0
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
This figure presents the cumulative magnitude beginning 20 trading days prior to the SEO issue date
before and after the Securities Offering Reform (SOR) Magnitude is calculated as the absolute value of
daily return net of the return on the CRSP value-weighted index
37
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity
offering (SEO)
20
18
Ag
gre
ga
te m
ag
nit
ud
e (p
ercen
t)
16
14
12
10
8
6
4
2
0
Trading days prior to SEO issue date
Pre-SOR Post-SOR
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
The figure plots three-day [-1+1] magnitude aggregated for all 8-K filing dates beginning 20 trading days
prior to the SEO issue date before and after the Securities Offering Reform (SOR) Three-day magnitude
is calculated as the cumulative three-day [-1+1] absolute value of daily returns net of the returns on the
CRSP value-weighted index
38
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Table 1 Sample distribution and offering characteristics
Panel A Distribution of SEOs by calendar year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 16 15 15 12 17 15 12 24 55 71 126
Non-forecasting 14 23 30 26 19 31 36 55 90 144 234
Full Sample 30 38 45 38 36 46 48 79 145 215 360
Panel B Offering characteristics for SEOs by calendar year
Full
Pre- Post- Sample
Variable 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Shelf (percent) 500 763 822 895 861 826 917 861 752 870 822
Registration days 621 1878 2136 2661 2455 1647 3926 3618 1960 2984 2571
Pre-issue days 34 37 35 26 18 38 20 11 33 20 26
SEO proceeds (millions of dollars) 2372 2232 3564 3400 2805 3508 6239 3235 2872 3937 3508
SEO size (percent) 106 137 127 123 114 140 98 131 123 124 123
Panel C Average number of 8-Ks filed during SEO fiscal year
Full
Pre- Post- Sample
Sample 2002 2003 2004 2005 2006 2007 2008 2009 SOR SOR Period
Forecasting 64 128 133 194 159 141 166 170 128 160 144 Non-forecasting 99 135 143 223 193 189 190 190 155 190 177
Full Sample 81 130 144 216 177 170 185 183 146 179 165
This table presents the distribution values for our full sample of 360 seasoned equity offerings (SEOs) between 2002 and 2009 The Securities
Offering Reform (SOR) became effective December 1 2005 Panels A and B are presented at the SEO level and Panel C is presented at the issuer-
year level We designate SEOs issued between January 1 2002 and November 30 2005 as pre-SOR and SEOs issued between December 1 2005
and December 31 2009 as post-SOR Three SEOs in the 2005 forecasting sample and three SEOs in the 2005 non-forecasting sample occur in
December 2005 These SEOs are included in the 2005 and post-SOR columns The final column presents the total number in Panel A and the
average values in Panels B and C for the full sample period The forecasting (non-forecasting) sample contains SEOs in which the firm has chosen
(not) to provide at least one annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue
date All variables are defined in the Appendix
39
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Market capitalization (billions of dollars)
Institutional ownership (percent)
522506
6180
601371
7262
888659
6709
850331
7500
298746
5856
478620
7144
Analyst following
Analyst dispersion (percent)
Litigation risk (percent)
Book-to-market
1039
1408
3172
038
1087
2218
2233
044
935
1446
2545
041
1011
797
2254
044
1102
1384
3556
036
1122
2919
2222
044
Loss (percent)
Firm risk (percent)
2897
252
3535
376
1636
273
1690
323
3667
280
4444
402
This table presents the mean values of descriptive statistics for our control variables The forecasting (non-forecasting) sample contains seasoned
equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or quarterly point or range management earnings forecast
(MEF) or both within one year before the SEO issue date The full sample contains both the forecasting and non-forecasting samples and
indicate that the difference in means is significant at the 1 5 and 10 level respectively using a standard two-tailed t-test All variables are
defined in the Appendix
40
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N= 215) (N = 55) (N = 71) (N = 90) (N = 144)
Panel A Occurrence of guidance disclosure during one month prior to SEO
Annual MEF 1034 1488 2727 4507
mdash mdash Quarterly MEF 1241 1163 3273 3531 mdash mdash All MEF 1724 1953 4545 5915 mdash mdash FWP NA 930 NA 1268 NA 764
Guidance 1724 2605 4545 6338 mdash 764
Panel B Frequency of disclosure during one month prior to SEO
Total disclosure 246 308
271 358
231 283
Annual MEF 012 020 031 059
mdash mdash Quarterly MEF 014 017 038 051 mdash mdash All MEF 026 036 069 110
mdash mdash
FWP NA 011 NA 013 NA 010
Guidance 026 047
069 123
mdash 010
Earnings announcement 047 048 044 044 049 051
8-K 173 213
158 192
182 223
Panel C Size of Form 8-K disclosure during one month prior to SEO
8-K size 254 597
297 830
229 481
8-K abnormal size 221 517
267 754 194 399
Panel D Distance between last disclosure and SEO issue date for prior one-month
Annual MEF mdash mdash 1473 1188 mdash mdash Quarterly MEF mdash mdash 1494 904
mdash mdash
All MEF mdash mdash 1456 1052
mdash mdash FWP NA 315 NA 278 NA 345
Guidance mdash mdash 1456 971
mdash mdash Earnings announcement 1440 1346 1329 1439 1500 1307
8-K 813 684 860 837 783 605
Closest disclosure 785 615
824 672 760 586
This table presents mean values for multiple forms of disclosure before and after SOR The forecasting (non-forecasting)
sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The
full sample contains both the forecasting and non-forecasting samples Panel D is for the subgroup of SEO firms with the
respective type of disclosure in the prior one-month period and indicate significance at the 1 5 and 10
level respectively using a standard two-tailed t-test NA = not applicable All variables are defined in the Appendix
41
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
Panel A Occurrence frequency and 8-K size during one-month period prior to SEO issue date
Variable
Occurrence of
annual MEF
Occurrence of
guidance
Frequency of
guidance
Frequency of
8-Ks
Frequency of
total disclosure
File size of
last 8-K
File size of
last 8-K
Constant
(1) -189
(2) -125
(3) -002
(4) 085
(5) 154
(6) -1216
(7) -1223
(005) (018) (098) (014) (005) (021) (020)
SOR 043
047
048
038
055
198
492
(009) (006) (004) (001) (001) (006) (004)
Ln(market capitalization) 010 -002 -002 014
010 151 173
(032) (083) (087) (009) (034) (013) (011)
Institutional ownership 110
113
107
-019 036 235 189
(004) (001) (001) (045) (030) (031) (039)
Residual analyst following -002 -006
-004
004
002 001 002
(046) (004) (005) (010) (046) (095) (091)
Analyst dispersion -474
076 038 -011 -014 515 497
(006) (019) (029) (051) (060) (034) (035)
Litigation risk -020 025 -002 -017 -010 -114 -106
(051) (041) (091) (032) (067) (036) (042)
Book-to-market 018 071
024 -015 -001 327 363
(047) (003) (010) (031) (098) (030) (027)
Loss -036 -068
-041
008 -027 -222
-220
(034) (007) (008) (062) (019) (004) (005)
Firm risk 156 391 041 159 340 3324 6409
(086) (064) (094) (065) (049) (024) (011)
Non-SEO 8-K size -690
(014)
Sample Forecasting Forecasting Forecasting Full Full Full Full
Regression Probit Probit OLS OLS OLS OLS OLS
Observations 126 126 126 360 360 358 358
Pseudo R2
0150 0120
R2
010 006 004 006 008
Adjusted R2
003 003 002 004 005
42
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Table 4 (continued)
Panel B Distance between last disclosure and SEO issue date by disclosure type
Variable
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Distance of
last guidance
Distance of
last 8-K
Distance of
closest disclosure
Constant
(1) 2456
(2) 1595
(3) 1647
(4) 10149
(5) 1684
(6) 1690
(000) (000) (000) (019) (008) (001)
SOR -563
-083 -106 -1693 -691
-372
(003) (033) (020) (027) (002) (004)
Ln(market capitalization) -112 -104
-102
509 017 -043
(018) (000) (000) (059) (088) (060)
Institutional ownership 350 082 002 -5322
206 -017
(046) (058) (099) (008) (059) (095)
Residual analyst following 001 -023
-018
299 -046
-029
(096) (001) (003) (021) (002) (006)
Analyst dispersion -162 -053 -035 -3270 -157 -156
(064) (060) (070) (018) (053) (041)
Litigation risk -087 020 027 -668 152 216
(070) (083) (076) (073) (062) (026)
Book-to-market 023 034 -031 -633 -372 042
(085) (069) (065) (048) (045) (073)
Loss -020 -096 -085 5399
164 226
(096) (035) (037) (005) (049) (025)
Firm risk -11517
-1358 -2299 -69596 -4750 -6360
(006) (052) (023) (011) (024) (004)
Restricted to one month Yes Yes Yes No No No
Regression OLS OLS OLS OLS OLS OLS
Observations 81 322 332 146 358 360
R2
018 006 006 010 006 005
Adjusted R2
008 003 004 004 003 003
In Panel A the forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one annual or
quarterly point or range management earnings forecast (MEF) or both within one year before the SEO issue date The full sample contains both the forecasting
and non-forecasting samples In Panel B Columns 1ndash3 (4ndash6) contain SEOs in which the firm provides a specified disclosure within one month (one year) before
the SEO issue date p-values in parentheses are based on robust standard errors clustered at the firm level and indicate significance at the 1 5 and
10 level respectively from a two-tailed test All variables are defined in the Appendix
43
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Table 5 Univariate comparison of error and bias in management earnings forecasts
(MEFs) before and after the Securities Offering Reform (SOR)
Panel A Last MEF during one month prior to SEO
Pre-SOR Post-SOR
Variable (N = 25) (N = 42) p-value
Error 243 072 000
Bias 133 023 009
Panel B First annual MEF during one year versus last annual MEF during one month prior to
SEO
Pre-SOR Post-SOR
Variable (N = 18) (N = 35) p-value
Error
First annual MEF 266 136 008
Last annual MEF 247 077 001
Difference -020 -059
039
Bias
First annual MEF 182 071 020
Last annual MEF 178 032 005
Difference -003 -040
046
Horizon
First annual MEF 33139 33837 085
Last annual MEF 21567 17494 020
Panel C First quarterly MEF during one year versus last quarterly MEF during one month prior
to SEO
Pre-SOR Post-SOR
Variable (N = 21) (N = 29) p-value
Error
First quarterly MEF 136 093 030
Last quarterly MEF 234 084 003
Difference 098 -009 006
Bias
First quarterly MEF 051 -027 014
Last quarterly MEF 106 -039 007
Difference 055 -012 031
Horizon
First quarterly MEF 4319 2917 017
Last quarterly MEF 3433 3066 069
Panel A presents the mean values of error and bias for seasoned equity offerings (SEOs) in which the
firm provides an MEF within one month before the SEO issue date Error and bias are annualized for
quarterly MEFs Values for annual MEFs are used if both annual and quarterly MEFs are provided
Panels B and C present the mean values of the first annual or quarterly MEF provided within one year
before the SEO issue date and the last annual or quarterly MEF provided within one month before the
SEO issue date In Columns 1ndash2 and indicate if difference in first and last error or bias is
significantly different from zero at the 1 5 and 10 level respectively using a standard two-tailed t-
test In Column 3 and indicate if difference in pre- and post-SOR means is significant at the 1
5 and 10 level respectively using a standard two-tailed t-test Variables are defined in the Appendix
44
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs)
before and after the Securities Offering Reform (SOR)
MEF error MEF error MEF bias MEF bias
Variable (1) (2) (3) (4)
Constant -015 061 -303 -200
(093) (067) (014) (020)
SOR -156
-104
-149
-136
(003) (003) (006) (001)
Ln(market capitalization) 025 001 043
027
(026) (097) (008) (016)
Institutional ownership -001 123 137 136
(099) (012) (033) (017)
Residual analyst following 009
-003 009 005
(009) (053) (015) (043)
Analyst dispersion -027 091 -231
-189
(081) (043) (005) (004)
Litigation risk -066 -087
-108
-103
(011) (001) (003) (001)
Book-to-market 156
054 189
072
(010) (021) (007) (012)
Loss -002 092 -038 114
(098) (025) (075) (020)
Firm risk -240 -934 293 177
(092) (041) (091) (089)
Horizon 046 072
-018 034
(053) (002) (084) (051)
Sample Forecasting Forecasting Forecasting Forecasting
Restricted to one month Yes No Yes No
Regression OLS OLS OLS OLS
Observations 67 126 67 126
R2
025 018 024 015
Adjusted R2
012 010 010 007
The forecasting sample in Columns 1 and 3 (2 and 4) contains seasoned equity offerings (SEOs) in which
the firm has chosen to provide at least one annual or quarterly point or range MEF or both within one
month (one year) before the SEO issue date Error and bias are based on the last annual or annualized
quarterly MEF provided during the specified time period p-values in parentheses are based on robust
standard errors clustered at the firm level and indicate significance at the 1 5 and 10 level
respectively from a two-tailed test Variables are defined in the Appendix
45
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering
Reform (SOR)
Panel A Pre- and post-SOR magnitude
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Magnitude [-200] 3756 5520 3650 4332 3822 6106
(000) (006) (000)
Magnitude [1+20] 2906 3918 2755 3117 2991 4313
(000) (018) (000)
Abnormal magnitude 855 1602 895 1216 830 1793
(000) (022) (000)
Panel B Pre-SOR magnitude and post-SOR magnitude by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Magnitude [-200]
(N = 145)
3756
(N = 109)
5831
(N = 106)
5201
(N = 101)
5936
(N = 114)
5151
Magnitude [1+20] 2906
(000)
3823
(000)
4016
(000)
3835
(000)
3992
Abnormal magnitude 855
(000)
2008
(000)
1185
(000)
2102
(000)
1160
(000) (022) (000) (024)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
management earnings forecast (MEF) within one year before the SEO issue date Magnitude is the cumulative absolute value of daily return net of
the CRSP value-weighted index return over the specified trading days with day 0 being the volume-based corrected SEO issue date (Corwin
2003) Abnormal magnitude is magnitude [-20 0] less magnitude [+1+20] Panel A tests for statistical differences between mean values of
magnitude for the stated pre- and post-SOR samples Panel B tests for statistical differences between mean values of magnitude for the stated post-
SOR subsample and the pre-SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an
8-K) within one week prior to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a
standard two-tailed t-test p-values from these tests are displayed in parentheses Variables are defined in the Appendix
46
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-
Table 8 Stock returns before and after the Securities Offering Reform (SOR)
Panel A Pre- and post-SOR return
Full sample Forecasting Non-forecasting
Pre-SOR Post-SOR Pre-SOR Post-SOR Pre-SOR Post-SOR
Variable (N = 145) (N = 215) (N = 55) (N = 71) (N = 90) (N = 144)
Return [-200] 245 503 200 072 272 716
(025) (057) (017)
Return [1+20] 160 031 118 -071 185 082
(030) (031) (053)
Abnormal return 085 472 082 143 087 634
(011) (082) (010)
Panel B Pre-SOR return and post-SOR return by disclosure within one week of the SEO issue date
Pre-SOR Post-SOR Post-SOR
Full sample Short distance disclosure Short distance 8-K
Yes No Yes No
Variable
Return [-200]
(N = 145)
245
(N = 109)
836
(N = 106)
161
(N = 101)
936
(N = 114)
120
(009) (072) (007) (059)
Return [1+20] 160 088 -027 143 -038
Abnormal return 085
(065)
748
(022)
188
(092)
793
(014)
188
(006) (072) (006) (070)
The forecasting (non-forecasting) sample contains seasoned equity offerings (SEOs) in which the firm has chosen (not) to provide at least one
annual or quarterly point or range management earnings forecast (MEF) or both within one year before the SEO iss ue date Return is the
cumulative raw daily returns over the specified number of trading days with day 0 being the volume -based corrected SEO issue date (Corwin
2003) Abnormal return is return [-200] less return [+1+20] Panel A tests for statistical differences between mean values of return for the stated
pre- and post-SOR samples Panel B tests for statistical differences between mean values of return for the stated post-SOR subsample and the pre-
SOR full sample Short distance disclosure (8-K) contains SEOs in which the firm provides any form of disclosure (an 8-K) within one week prior
to the SEO issue date and indicate significant differences at the 1 5 and 10 level respectively using a standard two-tailed t-test p-
values from these tests are displayed in parentheses Variables are defined in the Appendix
47
- Cover page
-
- DERA Working Paper Header
- Information Environment under SOR
-
- Information Environment under SOR Rev3
-
- 1 Introduction
- 2 Background and empirical questions
-
- 21 Before the Securities Offering Reform of 2005
- 22 After the Securities Offering Reform of 2005
- 23 Empirical tests
- 24 Capital formation period
-
- 3 Data and sample selection
-
- 31 Disclosures analyzed
- 32 Sample selection
-
- 4 Research design
-
- 41 Occurrence frequency 8-K size and distance
- 42 Management earnings forecast accuracy
- 43 Information magnitude and stock returns
- 44 Comparison with other studies
-
- 5 Results
-
- 51 Descriptive statistics
- 52 Results for occurrence frequency 8-K size and distance
- 53 Results for management earnings forecast accuracy
- 54 Results for information magnitude and stock returns
-
- 6 Conclusion
- Appendix Variable definitions and other key terms
- References
- Fig 1 Timeline of shelf registration and issue before and after the Securities Offering Reform (SOR)
- Fig 2 Cumulative magnitude over one month prior to seasoned equity offering (SEO)
- Fig 3 Aggregate magnitude for all 8-Ks filed during one month prior to seasoned equity offering (SEO)
- Table 1 Sample distribution and offering characteristics
- Table 2 Descriptive statistics before and after the Securities Offering Reform (SOR)
- Table 3 Univariate comparison of disclosure before and after the Securities Offering Reform (SOR)
- Table 4 Multivariate analysis of disclosure before and after the Securities Offering Reform (SOR)
- Table 5 Univariate comparison of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 6 Multivariate analysis of error and bias in management earnings forecasts (MEFs) before and after the Securities Offering Reform (SOR)
- Table 7 Magnitude of information reflected in absolute market-adjusted returns before and after the Securities Offering Reform (SOR)
- Table 8 Stock returns before and after the Securities Offering Reform (SOR)
-